Is Joerg Schilling the next Linus Torvalds?  Joerg has advantages that 
Linus never had.  25 years of Unix engineering expeince.  And a 
codebase battle hardened by professional Ivy league engineers.  Unlike 
Linux, SchilliX, Joerg's OpenSolaris distro will support the proper 
SCSI commands.  It won't be just a kernel.  

AIX is long in the tooth compared to Solaris 10.   AIX lacks DTrace, 
Zones and other goodies.  Linux doesn't have the stability or the 
features of Solaris either, but it was open source.

But not any longer, as OpenSolaris has been confirmed to be released 
in a weeks time from June 8th!  OpenSolaris, SchilliX, and other 
OpenSolaris distros go head to head with Linux.  I'll bet the 
superiour offering will kick some serious linux (and AIX) butt.

http://www.osdir.com/Article5883.phtml

"Ralph" <no@way.com> wrote in message news:uM-dna3Pg_ZdEzrfRVn-qw@rcn.net...
> Mike Cox wrote:
>
>> Is Joerg Schilling the next Linus Torvalds?  Joerg has advantages that
>> Linus never had.  25 years of Unix engineering expeince.  And a
>> codebase battle hardened by professional Ivy league engineers.  Unlike
>> Linux, SchilliX, Joerg's OpenSolaris distro will support the proper
>> SCSI commands.  It won't be just a kernel.
>>
>> AIX is long in the tooth compared to Solaris 10.   AIX lacks DTrace,
>> Zones and other goodies.  Linux doesn't have the stability or the
>> features of Solaris either, but it was open source.
>
> What "features"? Linux has had the capability to do "zones" for quite a
> while longer than Solaris and has much better hardware support. Linux is
> also FASTER.
>
>
>>
>> But not any longer, as OpenSolaris has been confirmed to be released
>> in a weeks time from June 8th!  OpenSolaris, SchilliX, and other
>> OpenSolaris distros go head to head with Linux.  I'll bet the
>> superiour offering will kick some serious linux (and AIX) butt.
>
> Your idea of what is "superior" is ludicrous.
>
>>
>> http://www.osdir.com/Article5883.phtml
>
> And will continue to die the way it and Sun has been for some time.
>
> One thing that might be nice, once Solaris becomes open source is that
> someone may fix the fucking stupid ^H BS!
>

This troll is full, nothing else to see everyone can move along now.
--
Rodrick R. Brown 

Mike Cox wrote:

> Is Joerg Schilling the next Linus Torvalds?  Joerg has advantages that
> Linus never had.  25 years of Unix engineering expeince.  And a
> codebase battle hardened by professional Ivy league engineers.  Unlike
> Linux, SchilliX, Joerg's OpenSolaris distro will support the proper
> SCSI commands.  It won't be just a kernel.
> 
> AIX is long in the tooth compared to Solaris 10.   AIX lacks DTrace,
> Zones and other goodies.  Linux doesn't have the stability or the
> features of Solaris either, but it was open source.

What "features"? Linux has had the capability to do "zones" for quite a
while longer than Solaris and has much better hardware support. Linux is
also FASTER. 


> 
> But not any longer, as OpenSolaris has been confirmed to be released
> in a weeks time from June 8th!  OpenSolaris, SchilliX, and other
> OpenSolaris distros go head to head with Linux.  I'll bet the
> superiour offering will kick some serious linux (and AIX) butt.

Your idea of what is "superior" is ludicrous. 

> 
> http://www.osdir.com/Article5883.phtml

And will continue to die the way it and Sun has been for some time. 

One thing that might be nice, once Solaris becomes open source is that
someone may fix the fucking stupid ^H BS! 

Mike Cox wrote:
> Is Joerg Schilling the next Linus Torvalds?
 Your're an idiot.

On 9 Jun 2005 01:42:11 +0200, Mike Cox <mikecoxlinux@yahoo.com> wrote
in message <<42a78253_2@x-privat.org>>:

> Is Joerg Schilling the next Linus Torvalds?

No.

> SchilliX, Joerg's OpenSolaris distro will support the proper 
> SCSI commands.

Wow.  SCSI commands.  Wow.

> It won't be just a kernel.

Then it won't do anything.

> But not any longer, as OpenSolaris has been confirmed to be released 
> in a weeks time from June 8th!

Wow.  Closed open source.  Wow.

> http://www.xxxxx.com/Article####.htm

Thanks, Oxtard.

On 2005-06-08, Mike Cox <mikecoxlinux@yahoo.com> wrote:

> OpenSolaris distros go head to head with Linux.  I'll bet the 
> superiour offering will kick some serious linux (and AIX) butt.

Hm, you seem to have misplaced your post. comp.os.solaris.advocacy
is -----> that way.

-- 
Jurjen Oskam

In comp.unix.solaris Mike Cox <mikecoxlinux@yahoo.com> wrote:
> But not any longer, as OpenSolaris has been confirmed to be released 
> in a weeks time from June 8th!  OpenSolaris, SchilliX, and other 

Confirmed? By who? The lady who works at the corner shop down the road?
I'll put _serious_ money on that date being wrong.

  Scott

Mike Cox poked his little head through the XP firewall and said:

> Is Joerg Schilling the next Linus Torvalds?  Joerg has advantages that 
> Linus never had.

comp.unix.solaris plonked

-- 
When all you have is a hammer, everything looks like a nail.

Mike Cox wrote:
> Is Joerg Schilling the next Linus Torvalds?  Joerg has advantages that
> Linus never had.  25 years of Unix engineering expeince.  And a
> codebase battle hardened by professional Ivy league engineers.  Unlike
> Linux, SchilliX, Joerg's OpenSolaris distro will support the proper
> SCSI commands.  It won't be just a kernel.
>
> AIX is long in the tooth compared to Solaris 10.   AIX lacks DTrace,
> Zones and other goodies.  Linux doesn't have the stability or the
> features of Solaris either, but it was open source.
>
> But not any longer, as OpenSolaris has been confirmed to be released
> in a weeks time from June 8th!  OpenSolaris, SchilliX, and other
> OpenSolaris distros go head to head with Linux.  I'll bet the
> superiour offering will kick some serious linux (and AIX) butt.
>
> http://www.osdir.com/Article5883.phtml


Hey Mike, get off the crack pipe dude. Have you ever actually admin'ed
both Solaris and AIX? Sun is in so much trouble they spent half their
cash ($4.1 billion, with a B), on an aging hardware company.(Storage
Tech) Another not so brilliant move, in a long line of not so brilliant
moves. Revenue and market share continue on the slippery slope to
nowhere.

Get a grip.

> Linux is also FASTER.


Not according to the benchmarks that Sun is publishing.


http://www.sun.com/2004-1012/feature/

http://www.sun.com/software/solaris/benchmarks.jsp

In comp.unix.solaris Lin?nut <=?iso-8859-1?Q?lin=F8nut?=@bone.com> wrote:
> Mike Cox poked his little head through the XP firewall and said:
> 
>> Is Joerg Schilling the next Linus Torvalds?  Joerg has advantages that 
>> Linus never had.
> 
> comp.unix.solaris plonked
 
Hey, don't blame us for Mike Cox! That guy is a cross-platform troll.

Mike Cox wrote:

> But not any longer, as OpenSolaris has been confirmed to be released
> in a weeks time from June 8th!  OpenSolaris, SchilliX, and other
> OpenSolaris distros go head to head with Linux.  I'll bet the
> superiour offering will kick some serious linux (and AIX) butt.


Is this is the unix distro with CDDL crap bordering fake
open source license that all of 1350 people downloaded? 

On Thu, 9 Jun 2005, 7 wrote:

> Is this is the unix distro with CDDL crap bordering fake
> open source license that all of 1350 people downloaded?

No, it's the enterpise-class UNIX distro licensed under the
OSI-approved CDDL open source license , which more than 1.5
million people have downloaded so far.

-- 
Rich Teer, SCNA, SCSA, OpenSolaris CAB member

President,
Rite Online Inc.

Voice: +1 (250) 979-1638
URL: http://www.rite-group.com/rich

Does it run on PowerPC/POWER ? Does it come with decent LVM ? MPIO ? If
answer on one of this is NO it's not even worse taking a look at.

vlad.zam@gmail.com writes:

> Does it run on PowerPC/POWER ? Does it come with decent LVM ? MPIO ?
> If answer on one of this is NO it's not even worse taking a look at.

There's just no pleasing some people ...

Joe

Rich Teer wrote:
> On Thu, 9 Jun 2005, 7 wrote:
> 
> 
>>Is this is the unix distro with CDDL crap bordering fake
>>open source license that all of 1350 people downloaded?
> 
> 
> No, it's the enterpise-class UNIX distro licensed under the
> OSI-approved CDDL open source license , which more than 1.5
> million people have downloaded so far.
> 
....or maybe 1349 other people had to try a dozen times to download 
Solaris because Sun's bandwidth sucked and their ISOs kept getting 
corrupted. Perhaps Sun could do with a torrent...

Nick Ballard <nrballardco@yahoo.com> writes in comp.unix.solaris:
|Rich Teer wrote:
|> On Thu, 9 Jun 2005, 7 wrote:
|> 
|> 
|>>Is this is the unix distro with CDDL crap bordering fake
|>>open source license that all of 1350 people downloaded?
|> 
|> 
|> No, it's the enterpise-class UNIX distro licensed under the
|> OSI-approved CDDL open source license , which more than 1.5
|> million people have downloaded so far.
|> 
|...or maybe 1349 other people had to try a dozen times to download 
|Solaris because Sun's bandwidth sucked and their ISOs kept getting 
|corrupted. 

The 1.5 million is not the number of downloads, but the number of 
computers people who downloaded it reported they were installing it on.

|Perhaps Sun could do with a torrent...

Perhaps...

-- 
________________________________________________________________________
Alan Coopersmith * alanc@alum.calberkeley.org * Alan.Coopersmith@Sun.COM
 http://www.csua.berkeley.edu/~alanc/   *   http://blogs.sun.com/alanc/
  Working for, but definitely not speaking for, Sun Microsystems, Inc.

vlad.zam@gmail.com writes in comp.unix.solaris:
|Does it run on PowerPC/POWER ? 

Not yet, but that's one of the projects already proposed and being
worked on by some of community members Sun brought into the pre-release
pilot program.

|Does it come with decent LVM ? 

Logical Volume Manager?  The Solaris Volume Manager is fairly decent for
many configs.  (It's been a decade since I've used the AIX or Veritas
Volume Managers, so unfortunately, I can't really compare to the current
versions.)

-- 
________________________________________________________________________
Alan Coopersmith * alanc@alum.calberkeley.org * Alan.Coopersmith@Sun.COM
 http://www.csua.berkeley.edu/~alanc/   *   http://blogs.sun.com/alanc/
  Working for, but definitely not speaking for, Sun Microsystems, Inc.

In article <d8b8ii$12si$1@agate.berkeley.edu>,
Alan Coopersmith  <alanc@alum.calberkeley.org> wrote:
>|Perhaps Sun could do with a torrent...
>
>Perhaps...

Anyone from Sun blogging about the network infrastructure preparations
for Solaris 10 update 1?

My impression is that SDLC won't be able to handle the load from current
Solaris 10 FCS users looking to upgrade plus those folks who haven't
tried S10 because they're waiting for additional features or felt that
S10 FCS was too close to the bleeding edge.

John
groenveld@acm.org  

Ralph wrote:

> What "features"? Linux has had the capability to do "zones" for quite a
> while longer than Solaris and has much better hardware support. Linux is
> also FASTER.

Linux is not faster; that's just FUD. Don't be such sore losers.

> Is this is the unix distro with CDDL crap bordering fake
> open source license that all of 1350 people downloaded?

No, it's a true SVR4, POSIX-compliant, backward and forward compatible,
enterprise-class UNIX being open sourced for the first time in history.

Up until now we had amateur crap like Linux released for free, there
was priactuically no choice, but now there'll be a true enterprise
class, PROFESSIONAL UNIX available for free.

Hmmm, which one should I pick:

Linux, an amateur OS written by amateurs for amateurs

or

(Open)Solaris, a professionally ENGINEERED operating ENVIRONMENT?

Gee, what a dilemma.

vlad.zam@gmail.com wrote:
> Does it run on PowerPC/POWER ?

It did, and it will again. Port is udnerway.

> Does it come with decent LVM ? MPIO ? If

It comes with one of the most powerful and easiest to use LVMs ever.

Mike Cox wrote:

 > professional Ivy league engineers

You say that like it's a good thing.

 > AIX is long in the tooth compared to Solaris 10.

You say that like it's a bad thing.

 > AIX lacks DTrace

Since DTrace is "open source", I guess it will just be a matter of 
someone porting it, right? Or does it use hidden interfaces like certain 
other vendors?

 > Zones

Let's not go over the whole Zones vs LPAR thing again. Let's just agree 
to disagree.

 > other goodies.

What "other goodies"? If they mattered to anyone, you would mention them 
by name.

 > But not any longer, as OpenSolaris has been confirmed to be released 
in a weeks time from June 8th!

By whom? A bloke you spoke to in a bar? This from the article quoted: 
"Sun initially said it would release OpenSolaris in 2004."

But we will wait and see.

On Friday 10 June 2005 02:08, tripivceta@hotmail.com
<tripivceta@hotmail.com>
(<1118390897.092707.292620@g43g2000cwa.googlegroups.com>) wrote:

> Up until now we had amateur crap like Linux released for free,

Is that an official statement of Credit Suisse?

> Gee, what a dilemma.

Google + Hotmail == Detritus.

Arkady Duntov wrote:
> Is that an official statement of Credit Suisse?

Absolutely not, nor does it have anything to do with CS.

Arkady Duntov <arkady-duntov@brotherhood.ua> writes:
>On Friday 10 June 2005 02:08, tripivceta@hotmail.com
><tripivceta@hotmail.com>
>(<1118390897.092707.292620@g43g2000cwa.googlegroups.com>) wrote:
>
>> Up until now we had amateur crap like Linux released for free,
>
>Is that an official statement of Credit Suisse?

This kind of behaviour is generally frowned upon.


-- 
       "The road to Paradise is through Intercourse."
        [email me at huge [at] huge [dot] org [dot] uk]

In comp.unix.aix Alan Coopersmith <alanc@alum.calberkeley.org> wrote:
> vlad.zam@gmail.com writes in comp.unix.solaris:
> |Does it run on PowerPC/POWER ? 
> 
> Not yet,

Not quite right. It should read "not anymore". There was a Solaris 2.5/
2.5.1 which ran on the 43P, 40P/Series 850 and the Series 800 Laptops.
Besides that, at least in the Solaris 8 sourcecode, there was still the
whole POWER/PPC stuff.

> |Does it come with decent LVM ? 
> 
> Logical Volume Manager?  The Solaris Volume Manager is fairly decent for
> many configs.  (It's been a decade since I've used the AIX or Veritas
> Volume Managers, so unfortunately, I can't really compare to the current
> versions.)

AIX and veritas are both more flexible if it comes to dealing with
raw devices. But with filesystems, SDS/ODS isn't that bad, either.




		Martin (administrating both RS/6000s and Suns since 1994)

-- 
Remember to always speak softly.
But carry a big stick!

begin  In <1118390897.092707.292620@g43g2000cwa.googlegroups.com>, on
06/10/2005
   at 01:08 AM, tripivceta@hotmail.com said:

>Linux, an amateur OS written by amateurs for amateurs

Amateurs? Companies like IBM have sunk quite a bit of money into code
contributed to it.

You want amateur? How about not even supporting logical volumes[1]?
Not *THAT* is amateur.

>(Open)Solaris, a professionally ENGINEERED operating ENVIRONMENT?

See above. Of course, that could change, and I'd be interested in
playing with Solaris if it did change.

[1] No, extended logical partitions are not an m$-only hack. They're
    industry standard, even with people who wouldn't be caught dead
    running m$ malware.

-- 
Shmuel (Seymour J.) Metz, SysProg and JOAT  <http://patriot.net/~shmuel>

Unsolicited bulk E-mail subject to legal action.  I reserve the
right to publicly post or ridicule any abusive E-mail.  Reply to
domain Patriot dot net user shmuel+news to contact me.  Do not
reply to spamtrap@library.lspace.org

On Fri, 10 Jun 2005, Martin Etteldorf wrote:

> In comp.unix.aix Alan Coopersmith <alanc@alum.calberkeley.org> wrote:
> > vlad.zam@gmail.com writes in comp.unix.solaris:
> > |Does it run on PowerPC/POWER ?
> >
> > Not yet,
>
> Not quite right. It should read "not anymore". There was a Solaris 2.5/

You're both correct!  There was a Solaris 2.5.1 for PPC, which was canned
rather quickly.  But there is also an OpenSolaris project to port it to
current PPC hardware.  AFAIK, the latter is not running natively on PPC *yet*.

-- 
Rich Teer, SCNA, SCSA, OpenSolaris CAB member

President,
Rite Online Inc.

Voice: +1 (250) 979-1638
URL: http://www.rite-group.com/rich

>From the hostile tone of all the respondents it appears that many of
these 'freedom of choice' advocates are afraid of competition.

Apparently they are all for 'freedom of choice' as long as that choice
is either Linux or Linux. You would think that they would be more
receptive to other freely distributed operating systems.

But when a recognized name brand, officialy certified commercial grade
Unix becomes freely available they all scoff at it. Makes you wonder
why.

lqualig@uku.co.uk wrote:
>>From the hostile tone of all the respondents it appears that many of
> these 'freedom of choice' advocates are afraid of competition.
> 
> Apparently they are all for 'freedom of choice' as long as that choice
> is either Linux or Linux. You would think that they would be more
> receptive to other freely distributed operating systems.
> 
> But when a recognized name brand, officialy certified commercial grade
> Unix becomes freely available they all scoff at it. Makes you wonder
> why.
> 

I for one am all for freedom of choice. Yes, I use Windows, that's my 
/choice/, for quite a bit of /desktop/ work. I also use Linux for quite 
a lot of /desktop/ work and *all* of my /server/ setups. That's also my 
choice. I /chose/ to do my research and found that a hammer isn't the 
best thing to thread a needle with. Although you may /choose/ to employ 
a crude percussive device in such a manner.

-- 
Cheers,

Jim

-begin sig-
Opinions expressed in this message may or may not be representative of 
the opinions of its author. You decide.
Linux is not /user-friendly/. It *is* user-friendly. It is not 
ignorant-friendly, nor is it idiot-friendly.

Web: 	http://www.dotware.co.uk
	http://www.dotware-entertainment.co.uk

This is a battle of wits, and it is clear you are unarmed.

-end sig-

Jim <james@the-computer-shop.co.uk> writes:

> I for one am all for freedom of choice.

[snip]

> -begin sig-
> Opinions expressed in this message may or may not be representative of the
> opinions of its author. You decide.

[snip]

I like that part of your .signature.

The rest of it is too long though. But it's your choice.

Otherwise I completely agree with you. 

Bye, Dragan
 
-- 
Dragan Cvetkovic, 

To be or not to be is true. G. Boole      No it isn't.  L. E. J. Brouwer

!!! Sender/From address is bogus. Use reply-to one !!!

lqualig@uku.co.uk writes:

>>From the hostile tone of all the respondents it appears that many of
> these 'freedom of choice' advocates are afraid of competition.
>
> Apparently they are all for 'freedom of choice' as long as that choice
> is either Linux or Linux. You would think that they would be more
> receptive to other freely distributed operating systems.
>
> But when a recognized name brand, officialy certified commercial grade
> Unix becomes freely available they all scoff at it. Makes you wonder
> why.

Here are a number of things I might hold against Solaris:

1- Recognized name brand
2- officially certified
3- commercial grade

Those are the kind of things that appeal to corporate types,
not people that appreciate fine operating systems.

Just my opinion.

> Here are a number of things I might hold against Solaris:
>
> 1- Recognized name brand
> 2- officially certified
> 3- commercial grade
>
>
> Those are the kind of things that appeal to corporate types,
> not people that appreciate fine operating systems.



Just because something is recognized, officially certified and
commercial grade is not a reason to dismiss the product.

Do you wear a name brand sneaker or some no-name brand from K-Mart? Do
you fly on name-brand airlines or some unknown "Sky Friend" brand? Do
you buy a name brand automobile or do you build your own kit-car with
unbranded parts? Do you go to a AMA certified doctor or some hack who
works out of his apartment?

So if you are willing to accept "name brands" for clothing and you
don't have a problem with an AMA board certified doctor then why should
you hold this against an OS?

Why would anyone insist on ignoring an OS because it is "commercial
grade" or achieved an official certification. I would think that the
quality of the OS would be the first and foremost factor in your
decision and not some "anti-establishment" mentality.

In article <1118391008.948867.100060@z14g2000cwz.googlegroups.com>,
	tripivceta@hotmail.com writes:
> 
> 
> vlad.zam@gmail.com wrote:
>> Does it run on PowerPC/POWER ?
> 
> It did, and it will again. Port is udnerway.

Ah, the once and future port.

-- 
mailto:rlhamil@smart.net  http://www.smart.net/~rlhamil

Lasik/PRK theme music:
    "In the Hall of the Mountain King", from "Peer Gynt"

lqualig@uku.co.uk writes:

>> Here are a number of things I might hold against Solaris:
>>
>> 1- Recognized name brand
>> 2- officially certified
>> 3- commercial grade
>>
>>
>> Those are the kind of things that appeal to corporate types,
>> not people that appreciate fine operating systems.
>
>
>
> Just because something is recognized, officially certified and
> commercial grade is not a reason to dismiss the product.
>
> Do you wear a name brand sneaker or some no-name brand from K-Mart?

No name, but I don't shop at K-Mart.

> Do you fly on name-brand airlines or some unknown "Sky Friend" brand?

I usually take the plane that's going where I'm going.

> Do you buy a name brand automobile or do you build your own kit-car with
> unbranded parts?

If I was still into fooling around with cars, I'd have no problem
with a kit-car.  Certainly with an OS, I very much appreciate being
able to mess with the parts.

> Do you go to a AMA certified doctor or some hack who
> works out of his apartment?

A name brand doctor?

> So if you are willing to accept "name brands" for clothing and you
> don't have a problem with an AMA board certified doctor then why should
> you hold this against an OS?

By that logic, I should be using Windows only.  The biggest name
brand of all.

> Why would anyone insist on ignoring an OS because it is "commercial
> grade" or achieved an official certification. I would think that the
> quality of the OS would be the first and foremost factor in your
> decision and not some "anti-establishment" mentality.

I think you're reading too much into my statement.  I did say
"might hold against".  I'm just pointing out that the criteria
cited did nothing for me.  I've always been attracted to the
underdog.  If those are the only reasons to use Solaris,
I'm not convinced.

Tell me the OS is reliable, fast, flexible, (as Solaris is),
and then I'll listen.

In my everyday work, I use Solaris a lot.  I used to use
Sparc/Solaris as my desktop system at work and at home.
Unfortunately, Sun decided a while back that they could
sell servers without being competitive on the desktop and
I finally gave in and switched to Linux for desktop use.

Now, it's hard to tell what they are up to but to make me
switch, they have to do _better_ than Mandrake to interest me.
We'll see.

Dan Espen wrote:
> lqualig@uku.co.uk writes:

>>But when a recognized name brand, officialy certified commercial grade
>>Unix becomes freely available they all scoff at it. Makes you wonder
>>why.

> Here are a number of things I might hold against Solaris:
> 
> 1- Recognized name brand

So I guess if you played acoustic guitar, you'd steer clear of Martin
because it's a recognized name brand?

   - Logan

I agree with 7.  lmao

On Fri, 10 Jun 2005 17:37:40 -0000,
Richard.L.Hamilton@mindwarp.smart.net (Richard L. Hamilton) wrote:

>In article <1118391008.948867.100060@z14g2000cwz.googlegroups.com>,
>	tripivceta@hotmail.com writes:
>> 
>> 
>> vlad.zam@gmail.com wrote:
>>> Does it run on PowerPC/POWER ?
>> 
>> It did, and it will again. Port is udnerway.
>
>Ah, the once and future port.

Assuming you are not just being a prick, you can check with Dennis
Clarke over at Blastwave about the PPC port.

-david-

On Fri, 10 Jun 2005 12:00:00 -0400, Dan Espen
<daneNO@SPAM.mk.telcordia.com> wrote:

>lqualig@uku.co.uk writes:
>
>>>From the hostile tone of all the respondents it appears that many of
>> these 'freedom of choice' advocates are afraid of competition.
>>
>> Apparently they are all for 'freedom of choice' as long as that choice
>> is either Linux or Linux. You would think that they would be more
>> receptive to other freely distributed operating systems.
>>
>> But when a recognized name brand, officialy certified commercial grade
>> Unix becomes freely available they all scoff at it. Makes you wonder
>> why.
>
>Here are a number of things I might hold against Solaris:
>
>1- Recognized name brand
>2- officially certified
>3- commercial grade
>
>Those are the kind of things that appeal to corporate types,
>not people that appreciate fine operating systems.
>
>Just my opinion.

Apples and oranges. Points 1, 2 and 3 may or may not have anything to
do with being a 'fine' operating system. Since you don't define
'fine', anyone is free to declare any operating system as 'fine'
regardless of points 1, 2, or 3.

-david-
This is a 'fine' posting, it's a pity you won't appreciate it.

David Haynes <david.haynes@nospam.sympatico.ca> writes:

> On Fri, 10 Jun 2005 12:00:00 -0400, Dan Espen
> <daneNO@SPAM.mk.telcordia.com> wrote:
>
>>lqualig@uku.co.uk writes:
>>
>>>>From the hostile tone of all the respondents it appears that many of
>>> these 'freedom of choice' advocates are afraid of competition.
>>>
>>> Apparently they are all for 'freedom of choice' as long as that choice
>>> is either Linux or Linux. You would think that they would be more
>>> receptive to other freely distributed operating systems.
>>>
>>> But when a recognized name brand, officialy certified commercial grade
>>> Unix becomes freely available they all scoff at it. Makes you wonder
>>> why.
>>
>>Here are a number of things I might hold against Solaris:
>>
>>1- Recognized name brand
>>2- officially certified
>>3- commercial grade
>>
>>Those are the kind of things that appeal to corporate types,
>>not people that appreciate fine operating systems.
>>
>>Just my opinion.
>
> Apples and oranges. Points 1, 2 and 3 may or may not have anything to
> do with being a 'fine' operating system. Since you don't define
> 'fine', anyone is free to declare any operating system as 'fine'
> regardless of points 1, 2, or 3.
>
> -david-
> This is a 'fine' posting, it's a pity you won't appreciate it.

Jeez, a lot of touchy people here.
(Well for me, here is C.U.S.)

I think my original post was clear enough.
Being a recognized name brand, buying certification,
calling something commercial, does nothing for me.
Especially in the context of Open Solaris vs. Linux.

For all you "name brand, officially certified" supporters,
guess what, Linux isn't for you.  Why waste your time even looking.

> For all you "name brand, officially certified"
> supporters, guess what, Linux isn't for you.
> Why waste your time even looking.


I don't believe that it's a matter of being a "name brand, officially
certified" supporter. I think the issue people have is where you said
that you would hold these things *against* Solaris.

So it's not a matter of people expecting anyone to use it because "it's
name brand... certified...etc." It's more of an issue with refusing to
even consider it because it happens to be "name brand, certified, etc."

BTW - After your original response I now understand what you meant. But
your choice of words did leave the wrong initial impression.

Dan Espen <daneNO@SPAM.mk.telcordia.com> writes:

> I think my original post was clear enough.
> Being a recognized name brand, buying certification,
> calling something commercial, does nothing for me.
> Especially in the context of Open Solaris vs. Linux.

Who says that Open Solaris is certified? It would need to go through the
certification process, the same way Solaris went.

Dragan

-- 
Dragan Cvetkovic, 

To be or not to be is true. G. Boole      No it isn't.  L. E. J. Brouwer

!!! Sender/From address is bogus. Use reply-to one !!!

On Friday 10 June 2005 14:41, Dragan Cvetkovic <me@privacy.net>
(<lmekbarll0.fsf@privacy.net>) wrote:

> Dan Espen <daneNO@SPAM.mk.telcordia.com> writes:
> 
> Who says that Open Solaris is certified?

http://www.opengroup.org/openbrand/register/

Arkady Duntov <arkady-duntov@brotherhood.ua> writes:

> On Friday 10 June 2005 14:41, Dragan Cvetkovic <me@privacy.net>
> (<lmekbarll0.fsf@privacy.net>) wrote:
>
>> Dan Espen <daneNO@SPAM.mk.telcordia.com> writes:
>> 
>> Who says that Open Solaris is certified?
>
> http://www.opengroup.org/openbrand/register/

That's Solaris as released by Sun (Solaris 8, Solaris 9 etc), not Open
Solaris, which hasn't been released yet, but will be soon.

Bye, Dragan


-- 
Dragan Cvetkovic, 

To be or not to be is true. G. Boole      No it isn't.  L. E. J. Brouwer

!!! Sender/From address is bogus. Use reply-to one !!!

Rich Teer wrote:
> On Thu, 9 Jun 2005, 7 wrote:
> 
> 
>>Is this is the unix distro with CDDL crap bordering fake
>>open source license that all of 1350 people downloaded?
> 
> 
> No, it's the enterpise-class UNIX distro licensed under the
> OSI-approved CDDL open source license , which more than 1.5
> million people have downloaded so far.
> 

I am drooling, waiting for OpenSolaris, and also a Sunfire
(when funds are available in July).  But I wish to mention
a few concerns (and I know that Sun listens here).  Note I
do NOT consider myself a Solaris-qualified Sys Admin, however
I must be pretty close!

I don't think the market is the same for OpenSolaris as
for Solaris.  OpenSolaris will be used by home users,
who are *not* enterprise-savvy IT specialists.  Having
a typical home user, even perhaps a Linux user, administer
a Solaris box is a bit like throwing a non-swimmer into
the deep end of the pool.  Ultra-simplified administrative
tools and user interfaces will be a must.  Total backward
compatibility with older Solaris won't be seen as important.

Given these assumptions, I'd like to hope that OpenSolaris:

   Will clean up duplicate/redundant admin tools and files,
   so that information exists in a single place.

   Have home user default configurations for everything, so
   the user can ignore (at least initially) networking configuration,
   security configuration, etc. (I'm talking post-install config,
   of course the installer must ask the user questions.)

   Be friendly to alternative boot loaders.  Detecting Grub/LILO,
   being able to boot easily from those as well as Windows boot
   loader, is important.  There will be multi-boot systems!
   (Historically the SunOS and Solaris boot loaders haven't worked
   and played well with others.)

   Will have many books writing about it, including academic ones
   suitable for adoption in the classroom for teaching System
   Administration.  These must have lots of philosophy and concepts,
   e.g., configuration planning and management, policy and procedure
   documentation, patch management techniques, and of course plenty
   of how-tos: web server, Java application server, clusters/load balancing,
   server-room planning, etc.  You'll also need some "OpenSolaris
   for Dummies" type of books, with the usual titles: OpenSolaris Bible,
   Unleashed, Definitive Guide to, in 24 hours, in 21 days, ...
   (docs.sun.com is fantastic but doesn't replace books I can read over
   lunch!)  I'd love to learn that Sun has hired some excellent/famous
   authors to flood the market with titles.
   (Without books your target audience will be too shy to try it, and with
   out hype-less (no marketing stuff) textbooks, it will be hard to adopt
   in the schools in lieu of Linux.)

   For ease of learning, some break with backward administrative compatibility
   is useful.  For example all related command can share a common prefix,
   "pkg" for all package management", "mod" for module management, etc.  Using
   a unified naming scheme is a radical suggestion, but is only that so
   please don't flame me!  But think of the ease of learning, of teaching, and
   of remembering.  Or of using the shell's auto-completion feature to
   discover commands!  (In this regard full-text man page keyword searching
   ought to be enhanced, using search-engine-like syntax.)  How often in
   this ng do you see a newbie question "how do I do foo?" and the answer
   is use the "bar" utility.  Well, it is hard to discover the utilities to
   use now because they don't use any naming scheme.

That's my two cents worth anyway!

-Wayne

On Fri, 10 Jun 2005 16:21:40 -0400, Dan Espen wrote:

> For all you "name brand, officially certified" supporters,
> guess what, Linux isn't for you.  Why waste your time even looking.

Novell isn't "name brand, officially certified"? Redhat isn't?

Liam Slider <liam@nospam.liamslider.com> writes:

> On Fri, 10 Jun 2005 16:21:40 -0400, Dan Espen wrote:
>
>> For all you "name brand, officially certified" supporters,
>> guess what, Linux isn't for you.  Why waste your time even looking.
>
> Novell isn't "name brand, officially certified"? Redhat isn't?
>

If by "officially certified" you mean UNIX03, UNIX98 or similar certified,
I don't see them in http://www.opengroup.org/openbrand/register/


-- 
Dragan Cvetkovic, 

To be or not to be is true. G. Boole      No it isn't.  L. E. J. Brouwer

!!! Sender/From address is bogus. Use reply-to one !!!

"Dragan Cvetkovic" <me@privacy.net> wrote in message 
news:lmekbarll0.fsf@privacy.net...
>
>> I think my original post was clear enough.
>> Being a recognized name brand, buying certification,
>> calling something commercial, does nothing for me.
>> Especially in the context of Open Solaris vs. Linux.
>
> Who says that Open Solaris is certified? It would need to go through the
> certification process, the same way Solaris went.
>

I don't believe that "Open Solaris" is an operating system. It is the name 
of the 'program' where Sun will freely distribute the Solaris OS. You can 
download the OS from Sun for free.

http://www.sun.com/software/solaris/get.jsp

What you are downloading and getting is the Solaris 10 operating system. Not 
something called Open Solaris.

In comp.unix.aix Larry Qualig <Lqualig_do_remove@uku.co.uk> wrote:
> "Dragan Cvetkovic" <me@privacy.net> wrote in message 
> news:lmekbarll0.fsf@privacy.net...
>>
>>> I think my original post was clear enough.
>>> Being a recognized name brand, buying certification,
>>> calling something commercial, does nothing for me.
>>> Especially in the context of Open Solaris vs. Linux.
>>
>> Who says that Open Solaris is certified? It would need to go through the
>> certification process, the same way Solaris went.
>>

> I don't believe that "Open Solaris" is an operating system. It is
> the name of the 'program' where Sun will freely distribute the
> Solaris OS. You can download the OS from Sun for free.

> http://www.sun.com/software/solaris/get.jsp

> What you are downloading and getting is the Solaris 10 operating
> system. Not something called Open Solaris.

To expand on that a little.

From http://opensolaris.org/faq/general_faq.html

   What is the difference between the OpenSolaris source code and the
   Solaris source code?

   The OpenSolaris program provides the core open source code base
   that you are free to modify and distribute under terms of the
   CDDL. The OpenSolaris binary that will be available on the website
   at launch in Q2 will be supported by the community rather than by
   Sun Services. The Solaris OS is Sun's distribution of the operating
   system: Sun-branded, Sun-maintained, and Sun-supported. Future
   distributions of the Solaris operating system will be built from
   the OpenSolaris source code.

and

   Will Sun be open sourcing all of the Solaris source code?

   We intend to open source as much of the code as possible. Some
   components (such as drivers) that are based on intellectual
   property from other companies will be offered as binary code only.

I get the impression that the OpenSolaris (kernel?) is to Solaris from
Sun rather like Darwin is to Mac OS-X from Apple.

rick jones
-- 
Wisdom Teeth are impacted, people are affected by the effects of events.
these opinions are mine, all mine; HP might not want them anyway... :)
feel free to post, OR email to rick.jones2 in hp.com but NOT BOTH...

"Larry Qualig" <Lqualig_do_remove@uku.co.uk> writes:

> "Dragan Cvetkovic" <me@privacy.net> wrote in message 
> news:lmekbarll0.fsf@privacy.net...
>>
>>> I think my original post was clear enough.
>>> Being a recognized name brand, buying certification,
>>> calling something commercial, does nothing for me.
>>> Especially in the context of Open Solaris vs. Linux.
>>
>> Who says that Open Solaris is certified? It would need to go through the
>> certification process, the same way Solaris went.
>>
>
> I don't believe that "Open Solaris" is an operating system. It is the name 
> of the 'program' where Sun will freely distribute the Solaris OS. You can 
> download the OS from Sun for free.
>
> http://www.sun.com/software/solaris/get.jsp
>
> What you are downloading and getting is the Solaris 10 operating system. Not 
> something called Open Solaris.

That is what you are getting from the above link. However, start instead
from http://www.opensolaris.org and do some reading. Open Solaris is going
to be available for download soon. 

If it helps, you can think in terms Fedora vs. RH, but it's not that simple
....

Dragan

-- 
Dragan Cvetkovic, 

To be or not to be is true. G. Boole      No it isn't.  L. E. J. Brouwer

!!! Sender/From address is bogus. Use reply-to one !!!

Dan Espen wrote:
> I think my original post was clear enough.
> Being a recognized name brand, buying certification,
> calling something commercial, does nothing for me.

Pffffft.

Your original post *was* clear.  You said that they were
"things I might hold against Solaris", not things that
do nothing for you.  Those are two different things.

You've said something stupid, and now you're trying to say
you didn't say what you said.  Which means now you've said
*two* things that are stupid.

(BTW, don't take that as a personal attack.  It's not meant
as one.  I just don't feel like letting a comment like that
pass without objection.)

   - Logan

Logan Shaw <lshaw-usenet@austin.rr.com> writes:
> (BTW, don't take that as a personal attack.  It's not meant
> as one.  I just don't feel like letting a comment like that
> pass without objection.)

Once a linguist, always a linguist, eh? :-)

Dragan

-- 
Dragan Cvetkovic, 

To be or not to be is true. G. Boole      No it isn't.  L. E. J. Brouwer

!!! Sender/From address is bogus. Use reply-to one !!!

"Dragan Cvetkovic" <me@privacy.net> wrote in message 
news:lmslzprb7h.fsf@privacy.net...
>
> If it helps, you can think in terms Fedora vs. RH, but it's not that 
> simple
> ...

It's not perfectly clear what the situation is and what he future direction 
of this is. That's why I carefully worded my original post to say "I don't 
believe that..."

But you're right - "it's not that simple." But then again, with Sun involved 
what were you expecting?

"Rick Jones" <rick.jones2@hp.com> wrote in message 
news:%6qqe.6926$8l2.5652@news.cpqcorp.net...

>   The OpenSolaris program provides the core open source code base
>   that you are free to modify and distribute under terms of the
>   CDDL. The OpenSolaris binary that will be available on the website
>   at launch in Q2 will be supported by the community rather than by
>   Sun Services. The Solaris OS is Sun's distribution of the operating
>   system: Sun-branded, Sun-maintained, and Sun-supported. Future
>   distributions of the Solaris operating system will be built from
>   the OpenSolaris source code.
>
>
> I get the impression that the OpenSolaris (kernel?) is to Solaris from
> Sun rather like Darwin is to Mac OS-X from Apple.
>


I get a different impression depending on what web-page at Sun I'm currently 
reading. If you go to OpenSolaris.org then it looks one way. But from 
Sun.com/Solaris it appears to be different. I doubt that OpenSolaris was 
written from scratch. It most likely originated from Solaris. But future 
distributions of Solaris will be based on OpenSolaris. Isn't this like being 
your own grandfather?

Logan Shaw <lshaw-usenet@austin.rr.com> writes:

> Dan Espen wrote:
>> I think my original post was clear enough.
>> Being a recognized name brand, buying certification,
>> calling something commercial, does nothing for me.
>
> Pffffft.
>
> Your original post *was* clear.  You said that they were
> "things I might hold against Solaris", not things that
> do nothing for you.  Those are two different things.

What do you make of the word "might" in that sentence?
Don't answer.  If I haven't expressed myself well,
well I'm not running for office.

> You've said something stupid, and now you're trying to say
> you didn't say what you said.  Which means now you've said
> *two* things that are stupid.

Only two?  I'm doing pretty well today.

> (BTW, don't take that as a personal attack.  It's not meant
> as one.  I just don't feel like letting a comment like that
> pass without objection.)

OK.

"Larry Qualig" <Lqualig_do_remove@uku.co.uk> writes:

> I get a different impression depending on what web-page at Sun I'm currently 
> reading. If you go to OpenSolaris.org then it looks one way. But from 
> Sun.com/Solaris it appears to be different. I doubt that OpenSolaris was 
> written from scratch. It most likely originated from Solaris. But future 
> distributions of Solaris will be based on OpenSolaris. Isn't this like being 
> your own grandfather?

Here is another analogy: Netscape vs Mozzila (sp?)

Dragan

-- 
Dragan Cvetkovic, 

To be or not to be is true. G. Boole      No it isn't.  L. E. J. Brouwer

!!! Sender/From address is bogus. Use reply-to one !!!

In article <r8rja1pd5i4s8f7b4tqb906ll23ugrfia0@4ax.com>,
	David Haynes <david.haynes@nospam.sympatico.ca> writes:
> On Fri, 10 Jun 2005 17:37:40 -0000,
> Richard.L.Hamilton@mindwarp.smart.net (Richard L. Hamilton) wrote:
> 
>>In article <1118391008.948867.100060@z14g2000cwz.googlegroups.com>,
>>	tripivceta@hotmail.com writes:
>>> 
>>> 
>>> vlad.zam@gmail.com wrote:
>>>> Does it run on PowerPC/POWER ?
>>> 
>>> It did, and it will again. Port is udnerway.
>>
>>Ah, the once and future port.
> 
> Assuming you are not just being a prick, you can check with Dennis
> Clarke over at Blastwave about the PPC port.
> 

Well I sure knew that already (also see www.blastware.org [same folks] as
far as that goes).  I was just showing of my strange sense of humor with
that line...

-- 
mailto:rlhamil@smart.net  http://www.smart.net/~rlhamil

Lasik/PRK theme music:
    "In the Hall of the Mountain King", from "Peer Gynt"

Logan Shaw wrote:
> Dan Espen wrote:
> 
>> lqualig@uku.co.uk writes:
> 
> 
>>> But when a recognized name brand, officialy certified commercial grade
>>> Unix becomes freely available they all scoff at it. Makes you wonder
>>> why.
> 
> 
>> Here are a number of things I might hold against Solaris:
>>
>> 1- Recognized name brand
> 
> 
> So I guess if you played acoustic guitar, you'd steer clear of Martin
> because it's a recognized name brand?

So is Ford.  And I steer way clear of their stuff.

Hey, here's another wrench!  How about Apply buy Sun?  At least we'd
get a good desktop with well supported desktop applications.  Oh yeah,
maybe they can create a better forums environment for forums.sun.com.

Poke, poke!

"GreyCloud" <cumulus@mist.com> wrote in message 
news:A9CdnVWzvbgi3jffRVn-vw@bresnan.com...
> Logan Shaw wrote:
>> Dan Espen wrote:
>>
>>> lqualig@uku.co.uk writes:
>>
>>
>>>> But when a recognized name brand, officialy certified commercial grade
>>>> Unix becomes freely available they all scoff at it. Makes you wonder
>>>> why.
>>
>>
>>> Here are a number of things I might hold against Solaris:
>>>
>>> 1- Recognized name brand
>>
>>
>> So I guess if you played acoustic guitar, you'd steer clear of Martin
>> because it's a recognized name brand?
>
> So is Ford.  And I steer way clear of their stuff.

BMW and Porsche are also name brands. Now you're between a rock and a hard 
place. You want a no-name brand top quality car. Doooh!!

In article <42a9b238$76$fuzhry+tra$mr2ice@news.patriot.net>,
Shmuel (Seymour J.) Metz <spamtrap@library.lspace.org.invalid> wrote:
>begin  In <1118390897.092707.292620@g43g2000cwa.googlegroups.com>, on
>06/10/2005
>   at 01:08 AM, tripivceta@hotmail.com said:
>
>>Linux, an amateur OS written by amateurs for amateurs
>
>Amateurs? Companies like IBM have sunk quite a bit of money into code
>contributed to it.

Still magnitudes less than Sun did spend on Solaris....

If IBM did have a self written OS that was competitive, they would
not have used Linux.

-- 
EMail:joerg@schily.isdn.cs.tu-berlin.de (home) J�rg Schilling D-13353 Berlin
      js@cs.tu-berlin.de		(uni)  
      schilling@fokus.fraunhofer.de	(work) Blog: http://schily.blogspot.com/
URL:  http://cdrecord.berlios.de/old/private/ ftp://ftp.berlios.de/pub/schily

Joerg Schilling wrote:
> In article <42a9b238$76$fuzhry+tra$mr2ice@news.patriot.net>,
> Shmuel (Seymour J.) Metz <spamtrap@library.lspace.org.invalid> wrote:
> 
>>begin  In <1118390897.092707.292620@g43g2000cwa.googlegroups.com>, on
>>06/10/2005
>>  at 01:08 AM, tripivceta@hotmail.com said:
>>
>>
>>>Linux, an amateur OS written by amateurs for amateurs
>>
>>Amateurs? Companies like IBM have sunk quite a bit of money into code
>>contributed to it.
> 
> 
> Still magnitudes less than Sun did spend on Solaris....
> 
> If IBM did have a self written OS that was competitive, they would
> not have used Linux.

They have a bunch of them on the zSeries, yet still ported Linux to it..

Mike Cox wrote:
> Is Joerg Schilling the next Linus Torvalds?  Joerg has advantages that
> Linus never had.  25 years of Unix engineering expeince.  And a
> codebase battle hardened by professional Ivy league engineers.  Unlike
> Linux, SchilliX, Joerg's OpenSolaris distro will support the proper
> SCSI commands.  It won't be just a kernel.
>
> AIX is long in the tooth compared to Solaris 10.   AIX lacks DTrace,
> Zones and other goodies.  Linux doesn't have the stability or the
> features of Solaris either, but it was open source.
>
> But not any longer, as OpenSolaris has been confirmed to be released
> in a weeks time from June 8th!  OpenSolaris, SchilliX, and other
> OpenSolaris distros go head to head with Linux.  I'll bet the
> superiour offering will kick some serious linux (and AIX) butt.
>
> http://www.osdir.com/Article5883.phtml

Does OpenSolaris' X server run faster? Does it support composite? Can
it run Doom 3 and UT 2004? Is ZFS faster than XFS?

In article <v0wqe.2589$xr.918@fed1read05>,
Kevin Bowling  <kevin.bowling.no.spam@removenospam.wans.net> wrote:

>> If IBM did have a self written OS that was competitive, they would
>> not have used Linux.
>
>They have a bunch of them on the zSeries, yet still ported Linux to it..

Having a bunch does not mean having at least one competitive one....

A confectioner who makes good fancy cakes does not need to take
cakes from another confectioner, garnish them with own goodies and 
-- 
EMail:joerg@schily.isdn.cs.tu-berlin.de (home) J�rg Schilling D-13353 Berlin
      js@cs.tu-berlin.de		(uni)  
      schilling@fokus.fraunhofer.de	(work) Blog: http://schily.blogspot.com/
URL:  http://cdrecord.berlios.de/old/private/ ftp://ftp.berlios.de/pub/schily

On Fri, 10 Jun 2005 01:08:17 -0700, tripivceta wrote:

> Up until now we had amateur crap like Linux released for free, there
> was priactuically no choice, but now there'll be a true enterprise
> class, PROFESSIONAL UNIX available for free.
> 
> Hmmm, which one should I pick:
> 
> Linux, an amateur OS written by amateurs for amateurs
> 
> or
> 
> (Open)Solaris, a professionally ENGINEERED operating ENVIRONMENT?

Lets see, we take away all OSS from solaris, and check what is left. I
don't think your environment would be that great anymore. Because clearly,
you are talking about and _environment_ (as stated above), which is much
more then just a kernel.

Linux started out as an amateur OS, sure, but it isn't anymore, and all
that was created in that amateur period has already been replaced by newer
code, and continues to be improved.

I pitty you, for thinking that anything comming from SUN programmers
_must_ be good. I am pro OSS, but at least i know not everything is of
high quality.

-- 
Don't let your mind wander -- it's too little to be let out alone.

On Thu, 09 Jun 2005 01:42:11 +0200, Mike Cox wrote:

> Is Joerg Schilling the next Linus Torvalds?

No.

Simply because Linus is the person who makes the Linux kernel to what it
is, and Joerg is just like a distro packager.

-- 
Don't let your mind wander -- it's too little to be let out alone.

In comp.unix.aix Joerg Schilling <js@cs.tu-berlin.de> wrote:

>>Amateurs? Companies like IBM have sunk quite a bit of money into code
>>contributed to it.
> 
> Still magnitudes less than Sun did spend on Solaris....
> 
> If IBM did have a self written OS that was competitive, they would
> not have used Linux.

Well then... What exactly makes you thinking AIX wouldn't compete
with Solaris? I know both since they first came out and I'd say that
any of them has its own advantages and disadvantages.
Please let us participate in your wisdom and tell us what makes Solaris
so much better than AIX.


Regards,

		Martin.
-- 
Remember to always speak softly.
But carry a big stick!

In article <d8enrl$h4k$1@schlund.de>,
Martin Etteldorf  <etteldor@mail.ru> wrote:
>In comp.unix.aix Joerg Schilling <js@cs.tu-berlin.de> wrote:
>
>>>Amateurs? Companies like IBM have sunk quite a bit of money into code
>>>contributed to it.
>> 
>> Still magnitudes less than Sun did spend on Solaris....
>> 
>> If IBM did have a self written OS that was competitive, they would
>> not have used Linux.
>
>Well then... What exactly makes you thinking AIX wouldn't compete
>with Solaris? I know both since they first came out and I'd say that
>any of them has its own advantages and disadvantages.
>Please let us participate in your wisdom and tell us what makes Solaris
>so much better than AIX.

If you did read my previous text, you would not have to ask here...

It's IBM that believes that AIX (and in special the OS on the 370 like 
mainframes) that is not competive. So if you like to know more about
the backgrund, you would have to ask IBM.


-- 
EMail:joerg@schily.isdn.cs.tu-berlin.de (home) J�rg Schilling D-13353 Berlin
      js@cs.tu-berlin.de		(uni)  
      schilling@fokus.fraunhofer.de	(work) Blog: http://schily.blogspot.com/
URL:  http://cdrecord.berlios.de/old/private/ ftp://ftp.berlios.de/pub/schily

>> If IBM did have a self written OS that was competitive, they would
>> not have used Linux.

"It's IBM that believes that AIX (and in special the OS on the 370 like

mainframes) that is not competive. So if you like to know more about
the backgrund, you would have to ask IBM. "

Just to clear some misunderstandings (not on behalf of IBM as i don't
work for them) -AIX was NOT developed by IBM. It was developed by Bull
SA of France.
 IBM considers AIX more competitive than Linux and sells it for a
higher price than Linux. They have adopted Linux in a big way because
they want to mask out the differences between their pseries and
mainframes. I mean, the support (or lack thereof) of an OS on a piece
of h/w can negatively affect the sales numbers of that h/w.
 To the best of my knowledhge, no version of Linux (incl the one that
IBM sells) can do as good a job of scaling in terms of SMP performance
on POWER architecture as AIX does.
As a matter of fact, that is one of the reasons why it costs more to
buy AIX.

regards
-kamal

On Fri, 10 Jun 2005 18:32:56 -0400, Dragan Cvetkovic wrote:

> Liam Slider <liam@nospam.liamslider.com> writes:
> 
>> On Fri, 10 Jun 2005 16:21:40 -0400, Dan Espen wrote:
>>
>>> For all you "name brand, officially certified" supporters,
>>> guess what, Linux isn't for you.  Why waste your time even looking.
>>
>> Novell isn't "name brand, officially certified"? Redhat isn't?
>>
> 
> If by "officially certified" you mean UNIX03, UNIX98 or similar certified,
> I don't see them in http://www.opengroup.org/openbrand/register/

That's because Linux isn't Unix.

In comp.os.linux.advocacy lqualig@uku.co.uk wrote:
>> Here are a number of things I might hold against Solaris:
>>
>> 1- Recognized name brand
>> 2- officially certified
>> 3- commercial grade
>>
>>
>> Those are the kind of things that appeal to corporate types,
>> not people that appreciate fine operating systems.
> 
> 
> 
> Just because something is recognized, officially certified and
> commercial grade is not a reason to dismiss the product.
> 

Businesses today focus on the best *deal*, not the best quality.
A corporate seal of approval usually means the product is cheap
enough -- nothing more. 

"Commercial Grade".  What does that mean exactly?  

"Recognized name brand"?  I would not be suprised if "Linux" has more
widespread name recognition than either "Solaris" or "Unix".

"Officially Certified".  Hmm.  "MCSE" springs to mind.


> Do you wear a name brand sneaker or some no-name brand from K-Mart? Do
> you fly on name-brand airlines or some unknown "Sky Friend" brand? Do
> you buy a name brand automobile or do you build your own kit-car with
> unbranded parts? Do you go to a AMA certified doctor or some hack who
> works out of his apartment?
> 

Most folks go down to the corner Quick-Med franchise and see the Green-Card
Wonder with a degree from Bermuda School of Medicine because he's one of
the "corporate-ceritified" (i.e., cheapo) doctors that their crappy HMO 
will authorize.


> So if you are willing to accept "name brands" for clothing and you
> don't have a problem with an AMA board certified doctor then why should
> you hold this against an OS?
> 


NAFTA == who cares about brands when they're all crap nowadays.
If you ever wondered what the hell happened to the quality of Levi's
denim, just take a look at where it comes from nowadays.


> Why would anyone insist on ignoring an OS because it is "commercial
> grade" or achieved an official certification. I would think that the
> quality of the OS would be the first and foremost factor in your
> decision and not some "anti-establishment" mentality.
> 

Brands don't mean anything anymore.  Show me any industry catering to 
the general public that is not driven more by product cost than by
product quality.

In comp.unix.aix Joerg Schilling <js@cs.tu-berlin.de> wrote:

>>Well then... What exactly makes you thinking AIX wouldn't compete
>>with Solaris? I know both since they first came out and I'd say that
>>any of them has its own advantages and disadvantages.
>>Please let us participate in your wisdom and tell us what makes Solaris
>>so much better than AIX.
> 
> If you did read my previous text, you would not have to ask here...

I did, actually.

> It's IBM that believes that AIX (and in special the OS on the 370 like 
> mainframes) that is not competive.

Methinks you don't know what you're talking about. All I'm reading
from you are vague statements -- "if IBM did have", "they would not"...
Maybe you should read
<http://www.ibm.com/servers/eserver/pseries/migration/toaix.html> and
<http://www.ibm.com/aix/> before posting such stuff the next time.
And don't forget to take a look at the redbooks.

It may be _your_ opinion that IBM thinks, their OSes aren't
competitive. I beg to differ, and maybe you'll change your
opinion, too after you've read the pages mentioned above.

That said, I do understand that it's easier to cultivate his
own prejudices than trying to see something from a different
angle.


Regards,


		Martin.
-- 
Remember to always speak softly.
But carry a big stick!

On 2005-06-10, tripivceta@hotmail.com <tripivceta@hotmail.com> wrote:
>> Is this is the unix distro with CDDL crap bordering fake
>> open source license that all of 1350 people downloaded?
>
> No, it's a true SVR4, POSIX-compliant, backward and forward compatible,
> enterprise-class UNIX being open sourced for the first time in history.
>
> Up until now we had amateur crap like Linux released for free, there
> was priactuically no choice, but now there'll be a true enterprise
> class, PROFESSIONAL UNIX available for free.
>
> Hmmm, which one should I pick:
>
> Linux, an amateur OS written by amateurs for amateurs

	Someone should tell Oracle that.

	They're dumping Solaris for it.

[deletia]

	Why run an OS that only scales to 100 cpus when I can run an OS
that scales to 1000?


-- 
	The best OS in the world is ultimately useless         |||
	if it is controlled by a Tramiel, Jobs or Gates.      / | \
                                                     

On 2005-06-11, Joerg Schilling <js@cs.tu-berlin.de> wrote:
> In article <42a9b238$76$fuzhry+tra$mr2ice@news.patriot.net>,
> Shmuel (Seymour J.) Metz <spamtrap@library.lspace.org.invalid> wrote:
>>begin  In <1118390897.092707.292620@g43g2000cwa.googlegroups.com>, on
>>06/10/2005
>>   at 01:08 AM, tripivceta@hotmail.com said:
>>
>>>Linux, an amateur OS written by amateurs for amateurs
>>
>>Amateurs? Companies like IBM have sunk quite a bit of money into code
>>contributed to it.
>
> Still magnitudes less than Sun did spend on Solaris....
>
> If IBM did have a self written OS that was competitive, they would
> not have used Linux.

	AIX is quite competitive. It also doesn't require quite so
many extra cost addons as Solaris does in order to be fully 
functional.

-- 
	The best OS in the world is ultimately useless         |||
	if it is controlled by a Tramiel, Jobs or Gates.      / | \
                                                     

On 2005-06-10, Dan Espen <daneNO@SPAM.mk.telcordia.com> wrote:
> lqualig@uku.co.uk writes:
>
>>>From the hostile tone of all the respondents it appears that many of
>> these 'freedom of choice' advocates are afraid of competition.
>>
>> Apparently they are all for 'freedom of choice' as long as that choice
>> is either Linux or Linux. You would think that they would be more
>> receptive to other freely distributed operating systems.
>>
>> But when a recognized name brand, officialy certified commercial grade
>> Unix becomes freely available they all scoff at it. Makes you wonder
>> why.
>
> Here are a number of things I might hold against Solaris:
>
> 1- Recognized name brand
> 2- officially certified
> 3- commercial grade

	4- historically bad x86 support.	
	5- 3rd party vendors abandoning it in favor of Linux.
	6- Incomplete compared to other RISC Unixen.
	7- Scales to fewer cpus than "hobbyist" competitors.
	8- runs on slowest RISC hardware in the business.

>
> Those are the kind of things that appeal to corporate types,
> not people that appreciate fine operating systems.
>
> Just my opinion.


-- 
	The best OS in the world is ultimately useless         |||
	if it is controlled by a Tramiel, Jobs or Gates.      / | \
                                                     

In article <ZY6dnbNh-OMmqDbfRVn-rg@comcast.com>,
	JEDIDIAH <jedi@nomad.mishnet> writes:
>>>
>>>Amateurs? Companies like IBM have sunk quite a bit of money into code
>>>contributed to it.
>>
>> Still magnitudes less than Sun did spend on Solaris....

In terms of contributions to open source software (not Linux
in particular), Sun is second only to Berkeley in terms of
the amount of software contributed. I don't know how far
down the list IBM comes.

>> If IBM did have a self written OS that was competitive, they would
>> not have used Linux.
> 
> 	AIX is quite competitive. It also doesn't require quite so
> many extra cost addons as Solaris does in order to be fully 
> functional.

I'm curious about what extra cost addons you are thinking of.

-- 
Andrew Gabriel

Larry Qualig wrote:
> "GreyCloud" <cumulus@mist.com> wrote in message 
> news:A9CdnVWzvbgi3jffRVn-vw@bresnan.com...
> 
>>Logan Shaw wrote:
>>
>>>Dan Espen wrote:
>>>
>>>
>>>>lqualig@uku.co.uk writes:
>>>
>>>
>>>>>But when a recognized name brand, officialy certified commercial grade
>>>>>Unix becomes freely available they all scoff at it. Makes you wonder
>>>>>why.
>>>
>>>
>>>>Here are a number of things I might hold against Solaris:
>>>>
>>>>1- Recognized name brand
>>>
>>>
>>>So I guess if you played acoustic guitar, you'd steer clear of Martin
>>>because it's a recognized name brand?
>>
>>So is Ford.  And I steer way clear of their stuff.
> 
> 
> BMW and Porsche are also name brands. Now you're between a rock and a hard 
> place. You want a no-name brand top quality car. Doooh!!
> 

I place my bets on Toyota.  A good name brand of high quality and high 
reliability.

On 2005-06-11, Andrew Gabriel <andrew@cucumber.demon.co.uk> wrote:
> In article <ZY6dnbNh-OMmqDbfRVn-rg@comcast.com>,
> 	JEDIDIAH <jedi@nomad.mishnet> writes:
>>>>
>>>>Amateurs? Companies like IBM have sunk quite a bit of money into code
>>>>contributed to it.
>>>
>>> Still magnitudes less than Sun did spend on Solaris....
>
> In terms of contributions to open source software (not Linux
> in particular), Sun is second only to Berkeley in terms of
> the amount of software contributed. I don't know how far
> down the list IBM comes.
>
>>> If IBM did have a self written OS that was competitive, they would
>>> not have used Linux.
>> 
>> 	AIX is quite competitive. It also doesn't require quite so
>> many extra cost addons as Solaris does in order to be fully 
>> functional.
>
> I'm curious about what extra cost addons you are thinking of.
>

	Decent LVM & clusterware (of any sort) for a start.

-- 
	The best OS in the world is ultimately useless         |||
	if it is controlled by a Tramiel, Jobs or Gates.      / | \
                                                     

On Saturday 11 June 2005 09:18, Liam Slider <liam@nospam.liamslider.com>
(<pan.2005.06.11.15.16.21.94997@nospam.liamslider.com>) wrote:

> On Fri, 10 Jun 2005 18:32:56 -0400, Dragan Cvetkovic wrote:
> 
>> If by "officially certified" you mean UNIX03, UNIX98 or similar
>> certified, I don't see them in
>> http://www.opengroup.org/openbrand/register/
> 
> That's because Linux isn't Unix.

Apparently OpenSolaris isn't Unix either.

Nick Ballard wrote:
> ...or maybe 1349 other people had to try a dozen times to download 
> Solaris because Sun's bandwidth sucked and their ISOs kept getting 
> corrupted. Perhaps Sun could do with a torrent...

Perhaps, but wouldn't that just be another excuse for COLA/Linux people 
to accuse Sun being a bad, evil, twisted corporation taking advantage of 
the idealistic/good Freeware/Linux community?

Shmuel (Seymour J.) Metz wrote:
>>Linux, an amateur OS written by amateurs for amateurs
> 
> Amateurs? Companies like IBM have sunk quite a bit of money into code
> contributed to it.

Of course they have, together with HP and SGI, and they will continue to 
do so.

You see, these companies are betting that Linux will eventually grow in 
popularity (over a span of several years) to the volumes of Windows. 
They are betting that this will make their HW once again marketable and 
increase their sales volume.

Effectively they are throwing years of advanced engineering and billions 
of US dollars sunk into their operating environments, because they were 
unsuccessful in effectively marketing their product(s) in the post 
dot-com era.

You could also call it crawling on all fours and groveling.

Rather than firing their inept marketing goons and Microsoft prepaid 
CEOs, they're choosing the path of the least resistence.

However, all these facts do not change the one fact, and that is that 
Linux in its heart is a poorly hacked-together attempt of a UNIX like OS 
  by people THAT DO NOT TRULY UNDERSTAND UNIX.

You can know every command there is to know in UNIX, but that does not 
an engineer make!

> You want amateur? How about not even supporting logical volumes[1]?
> Not *THAT* is amateur.

What are you referring to?
If you are targeting Solaris, I must dissapoint you: I've worked with 
many LVMs over the years, from IRIX6.5 disk plexes/XLV to HP-UX LVM to 
Linux MDs and LVM to Solaris SVM, and I can with no qualms say that 
Solaris' LVM is by far the easiest logical volume manager of them all; 
and it's also very flexible.

It would appear that you're assuming that anything that doesn't use PEs 
and LEs isn't a volume manager proper. However, I have news for you: an 
LVM does not have to be designed around PEs and LEs. It's just a 
different philosophy, but the end result is the same.

> [1] No, extended logical partitions are not an m$-only hack. They're
>     industry standard, even with people who wouldn't be caught dead
>     running m$ malware.

Again, what are you referring to? Solaris supports the so called "soft 
partitions", allowing up to 128 of those.

Joerg Schilling wrote:

> If IBM did have a self written OS that was competitive, they would
> not have used Linux.

Correct.

begin  virus.txt.scr UNIX admin wrote:

> Nick Ballard wrote:
>> ...or maybe 1349 other people had to try a dozen times to download
>> Solaris because Sun's bandwidth sucked and their ISOs kept getting
>> corrupted. Perhaps Sun could do with a torrent...
> 
> Perhaps, but wouldn't that just be another excuse for COLA/Linux people
> to accuse Sun being a bad, evil, twisted corporation taking advantage of
> the idealistic/good Freeware/Linux community?

No. In geberal Cola/linux people are not interested in Solaris
They already have something better
-- 
A fool-proof method for sculpting an elephant:
first, get a huge block of marble; then you chip
away everything that doesn't look like an elephant.

Kevin Bowling wrote:

> They have a bunch of them on the zSeries, yet still ported Linux to it..

That's because IBM assumed (and correctly so) that a mainstream OS like 
Linux would make their aging mainframes more marketable.

The only problem is, the price of the aging mainframes didn't drop 
sufficiently, so IBM isn't seeing nearly the volumes sold that they'd 
hoped to see.

It turns out IBM learned a hard lesson: not all that shines is gold. 
Just because it's Linux doesn't make it an instant hit.

Martin Etteldorf wrote:

> Well then... What exactly makes you thinking AIX wouldn't compete
> with Solaris? I know both since they first came out and I'd say that
> any of them has its own advantages and disadvantages.
> Please let us participate in your wisdom and tell us what makes Solaris
> so much better than AIX.

Solaris is way more modern and supports way more hardware. Unlike AIX, 
it kept up with the times and it also sprung to two new platforms 
besides UltraSPARC/UPA/PCI, namely x86 and x64.

Martin Etteldorf wrote:

> It may be _your_ opinion that IBM thinks, their OSes aren't
> competitive. I beg to differ, and maybe you'll change your
> opinion, too after you've read the pages mentioned above.
> 
> That said, I do understand that it's easier to cultivate his
> own prejudices than trying to see something from a different
> angle.

Well, let's introduce new perspectives then, shall we?

I have no qualms about running AIX, in fact, I'd like to delve even more 
in depth than I already have.

Therefore, I started looking around for a pSeries system to buy for home 
use, and learn even more about AIX. The only stipulation was that it had 
to be 19" rack mountable.

It turns out that not only is the pSeries platform cost prohibitive (the 
cheapest model was close to $3000 USD) but on top of that, I'd have to 
shell out the dough for an AIX license to boot, which is also not cheap, 
to say the least.

Meanwhile, Ultra Enterprise450, 220R, 250R, and 280Rs are selling for 
$400-$600 USD on ebay, and I can get Solaris just with a few clicks, 
hassle free.

There goes my idea of doing more AIX. Thanks IBM, nice job!

JEDIDIAH wrote:

> 	AIX is quite competitive. It also doesn't require quite so
> many extra cost addons as Solaris does in order to be fully 
> functional.

No FUD please. Solaris is fully functional out-of-the-box, and even 
advanced tools like SVM come for free with the OS out-of-the-box.

dirk dierickx wrote:
> Lets see, we take away all OSS from solaris, and check what is left. I
> don't think your environment would be that great anymore. Because clearly,
> you are talking about and _environment_ (as stated above), which is much
> more then just a kernel.

Then you'd have to take away Solaris itself, because it too is OSS.

> Linux started out as an amateur OS, sure, but it isn't anymore, and all
> that was created in that amateur period has already been replaced by newer
> code, and continues to be improved.

As long as the Linux kernel developers themselves are claiming that 
Linux will not have ABI/API and driver forward/backward compatibility, 
Linux will not be able to even come close to being professional.

Let us not forget, that just because Linux is used by amateurs who act 
like they are professionals, that does not automatically make it a 
professional product.

And just because someone lands a job in IT does not make them an IT 
professional, which further reinforces the above point.

I'll give you a concrete exmaple:
I few weeks ago I had been interviewing for a potential job at a 
company. The "UNIX guys" there (all except one) were convinced that 
eventually they will be moving away from AIX to Linux, "because Linux is 
cheaper".
As soon as I heard that, it became crystal clear to me that these guys 
are amateurs. This had been further reinforced by one of the guys saying 
they are in the process of migrating from BIND to tinyDNS. Game over.

> I pitty you, for thinking that anything comming from SUN programmers
> _must_ be good. I am pro OSS, but at least i know not everything is of
> high quality.

Sun engineers are a group of highly skilled professional ENGINEERS who 
take great pride in the work they do. And they do a phenomenal job at 
ENGINEERING a PRODUCT.

If you know any better, you can tell just by looking at the work they 
did that these guys really understand UNIX and ENGINEERING.

JEDIDIAH wrote:
> 	Someone should tell Oracle that.

Someone did tell Oracle that, and Oracle apparently listened, since 
Solaris10 x86/x64 is now officially supported.

BTW, Larry will certify any platform for his product if he believes will 
bring in more profit. There's nothing personal in his decisions.
Having stated that, if Larry certified Solaris, it would imply that he 
deems it a profitable platform.


> 	They're dumping Solaris for it.

See above.

> 	Why run an OS that only scales to 100 cpus when I can run an OS
> that scales to 1000?

IRIX scales to thousands of CPUs, and so does Solaris. Linux does not, 
unless you use software clustering, which is not true CPU scaling.

lqualig@uku.co.uk wrote:
> Apparently they are all for 'freedom of choice' as long as that choice
> is either Linux or Linux. You would think that they would be more
> receptive to other freely distributed operating systems.

You forgot the mention the 3rd ALTERNATIVE: Linux.

In article <d8h2tn$gdm$03$1@news.t-online.com>,
Peter =?UTF-8?B?S8O2aGxtYW5u?=  <Peter.Koehlmann@t-online.de> wrote:

>No. In geberal Cola/linux people are not interested in Solaris
>They already have something better

Could you enlighten us what they use instead of Linux or Solaris?

-- 
EMail:joerg@schily.isdn.cs.tu-berlin.de (home) J�rg Schilling D-13353 Berlin
      js@cs.tu-berlin.de		(uni)  
      schilling@fokus.fraunhofer.de	(work) Blog: http://schily.blogspot.com/
URL:  http://cdrecord.berlios.de/old/private/ ftp://ftp.berlios.de/pub/schily

Dan Espen wrote:
> Here are a number of things I might hold against Solaris:
> 
> 1- Recognized name brand
> 2- officially certified
> 3- commercial grade
> 
> Those are the kind of things that appeal to corporate types,
> not people that appreciate fine operating systems.

If you are implying that Linux is a fine operating system, I do emplore 
you to delve deeper into operating systems, so that one day you may be 
able to tell the difference. Knowledge is power, as they say.

You do realize that the rest of your remarks against Solaris are not 
based on technical merit, don't you?

Dan Espen wrote:

> If I was still into fooling around with cars, I'd have no problem
> with a kit-car.  Certainly with an OS, I very much appreciate being
> able to mess with the parts.

Then why don't you start writing DDI/DDK drivers for Solaris? That's an 
excellent way to start "messing with the parts", as it offers some deep 
insights into Solaris internals.

Larry Qualig wrote:

> BMW and Porsche are also name brands. Now you're between a rock and a hard 
> place. You want a no-name brand top quality car. Doooh!!

This is OT, but before I abandon it as such, as I happen to work on cars 
for years: BMW is not a top quality car. Piss on any car that you can't 
even change oil on without having to reset the computer. That's no 
quality, that's overpriced bullshit with a name tag on it.

GreyCloud wrote:
> I place my bets on Toyota.  A good name brand of high quality and high 
> reliability.

I'm with you, although I don't own a Toyota.

Dan Espen wrote:

> For all you "name brand, officially certified" supporters,
> guess what, Linux isn't for you.  Why waste your time even looking.

I didn't look, I had to do it for a living. It put the bread on the 
table. Working with Linux day to day was one of the worst jobs I ever 
held; I was glad when I finally quit and went to work with Solaris again.

On Sun, 12 Jun 2005 13:04:23 +0200, UNIX admin wrote:

> IRIX scales to thousands of CPUs, and so does Solaris. Linux does not, 
> unless you use software clustering, which is not true CPU scaling.

except SGI is doing it too with linux, i think their biggest machine sold
was a 512 cpu 'box'.

-- 
Don't let your mind wander -- it's too little to be let out alone.

Rick Jones wrote:

> I get the impression that the OpenSolaris (kernel?) is to Solaris from
> Sun rather like Darwin is to Mac OS-X from Apple.

It's quite simple, actually: OpenSolaris will be like RedHat's Fedora 
core in terms of product placement.

On Sun, 12 Jun 2005 13:04:23 +0200, UNIX admin wrote:

> Someone did tell Oracle that, and Oracle apparently listened, since 
> Solaris10 x86/x64 is now officially supported.

sure, but the difference is that within oracle internally, they are
dumping solaris in favour of linux.

-- 
Don't let your mind wander -- it's too little to be let out alone.

Liam Slider wrote:

> Novell isn't "name brand, officially certified"? Redhat isn't?

Not in terms of UNIX. As it stands right now, no Linux distro/kernel is 
able to be UNIX certified. For that to happen, Linux would have to meet 
certain standards like POSIX and/or SysV, which it is currently unable 
to do, and that purely on technical merit.

The ironic thing is, if Linux would meet SVR4/POSIX standards, it would 
become a real high quality product, most likely even a professional 
operating environment.

Arkady Duntov wrote:

> Apparently OpenSolaris isn't Unix either.

Apparently. However, you're playing dumb, and you know it.
Since the first release of OpenSolaris will be a practically 1:1 copy of 
Solaris proper, and Solaris is UNIX, it therefore follows that 
OpenSolaris is UNIX.

Just to make sure you don't play dumb any more:

if A = B and B = C it follows that A = C, or,
if A = C and A = B then B = C.

You don't fool me. I know for a fact that the former Soviet Union had a 
very strong math curriculum.

JEDIDIAH wrote:

> 	4- historically bad x86 support.	

Not any more. Next!

> 	5- 3rd party vendors abandoning it in favor of Linux.

Not any more either: RealTek supports it out of the box, so does intel, 
so does nVidia, etcetera. Next!

> 	6- Incomplete compared to other RISC Unixen.

What exactly do you find incomplete?

> 	7- Scales to fewer cpus than "hobbyist" competitors.

Solaris easily runs on 1024 CPUs WITHOUT CLUSTERING. Next!

> 	8- runs on slowest RISC hardware in the business.

Tried UltraSPARC IV lately? Next!

And, it is irrelevant whether Solaris runs on SPARC/RISC or not, since 
it also runs on Pentium4 and Opteron, which are both CISC/RISC hybrids.

In comp.unix.aix UNIX admin <tripivceta@hotmail.com> wrote:
> It turns out that not only is the pSeries platform cost prohibitive (the 
> cheapest model was close to $3000 USD)

Well, I'd say it's reasonably priced. A comparably powerful x86 or
sparc machine won't be cheaper.

> but on top of that, I'd have to 
> shell out the dough for an AIX license to boot, which is also not cheap, 
> to say the least.

Nonsense. Pay 35 bucks for the Media kit. Any RS/6000 comes with a
valid AIX license.

> Meanwhile, Ultra Enterprise450, 220R, 250R, and 280Rs are selling for 
> $400-$600 USD on ebay,

You're comparing a new pSeries machine with an obsolete Ultra-II
machine? Get a life.

> and I can get Solaris just with a few clicks, 
> hassle free.

I wouldn't say it's too much of a hassle calling IBM or ordering
an AIX media kit online. Just one phonecall or just "a few clicks".

> There goes my idea of doing more AIX. Thanks IBM, nice job!

Go buy a 44P or alike. They are more comparable to an old U-II
machine in both price and performance.
But then, why am I talking to a troll anyway?



		Martin.
-- 
Remember to always speak softly.
But carry a big stick!

In comp.unix.aix UNIX admin <tripivceta@hotmail.com> wrote:
> Martin Etteldorf wrote:
> 
>> Well then... What exactly makes you thinking AIX wouldn't compete
>> with Solaris? I know both since they first came out and I'd say that
>> any of them has its own advantages and disadvantages.
>> Please let us participate in your wisdom and tell us what makes Solaris
>> so much better than AIX.
> 
> Solaris is way more modern [...]

Sounds interesting. Tell us more about this.


		Martin.

-- 
Remember to always speak softly.
But carry a big stick!

> Does OpenSolaris' X server run faster? Does it support composite?

It runs X.org, so it at least runs at the same speed or faster.

> Can it run Doom 3 and UT 2004?

It sure can, all you have to do is port those to Solaris.

 > Is ZFS faster than XFS?

I don't know that yet, besides, ZFS seems to be an LVM and an FS fused 
together, whereas XFS is just an FS. So it's like comparing oranges to 
apples.

On Sun, 12 Jun 2005 12:59:33 +0200, UNIX admin wrote:

> dirk dierickx wrote:
>> Lets see, we take away all OSS from solaris, and check what is left. I
>> don't think your environment would be that great anymore. Because clearly,
>> you are talking about and _environment_ (as stated above), which is much
>> more then just a kernel.
> 
> Then you'd have to take away Solaris itself, because it too is OSS.

ok, will that mean that starting from next week solaris will be a piece of
crap? because you know, those same monkeys messing up linux could do the
same to solaris? what are you going to use a year from now, when solaris
is reduced to a pile of smelling amateur poo?

anyway, you are avoiding my question! your view is that linux is badly
written, mostly because, i don't know, you have this idea the people
building it have no clue. i figure the same goes for the userland utils?
so if we take the OSS userland utils away from solaris, what will be left
of your precious UNIX? would any unix still be the same without all the
OSS tools?

> I'll give you a concrete exmaple:
> I few weeks ago I had been interviewing for a potential job at a 
> company. The "UNIX guys" there (all except one) were convinced that 
> eventually they will be moving away from AIX to Linux, "because Linux is 
> cheaper".

so, without further investigation, you decide that because they say it is
cheaper, they have no clue what they are doing? perhaps for their
individual case it _is_ the major advantage over others. in your thinking
you would turn down a job at google because they use linux for the same
reason (cheap).

>> I pitty you, for thinking that anything comming from SUN programmers
>> _must_ be good. I am pro OSS, but at least i know not everything is of
>> high quality.
> 
> Sun engineers are a group of highly skilled professional ENGINEERS who
> take great pride in the work they do. And they do a phenomenal job at
> ENGINEERING a PRODUCT.

i'm sure they are proud, still the product contains bugs just like any
other it is not perfect even though they are 'skilled profissional
ENGINEERS'. the linux coders are proud as well, many of them work at hp,
ibm, sgi, yes even sun, nasa, ... and their code is visible to _anybody_
wishing to look at it.

> If you know any better, you can tell just by looking at the work they
> did that these guys really understand UNIX and ENGINEERING.

but linux is not a unix, it is modeled after it, but it is something on
its own.

-- 
Don't let your mind wander -- it's too little to be let out alone.

Peter K�hlmann wrote:

> No. In geberal Cola/linux people are not interested in Solaris
> They already have something better

Or so they believe. But believing unfortunately isn't facts.

Martin Etteldorf wrote:
> Well, I'd say it's reasonably priced. A comparably powerful x86 or
> sparc machine won't be cheaper.

I just built a 1U/19", 3.2GHz, 4GB RAM, 2x160GB, 2x1Gb Solaris10 server 
based on P4/x86 for approx. $1800 USD or less.

If you can point me to a link that would sells a pSeries 1U/19" with the 
latest AIX 5.x server for that amount of money, I will be paying CASH, 
right away, NO QUESTIONS ASKED.

> Nonsense. Pay 35 bucks for the Media kit. Any RS/6000 comes with a
> valid AIX license.

That's something. Could you post a link please?

> You're comparing a new pSeries machine with an obsolete Ultra-II
> machine? Get a life.

Since when is B50 new? You really are taking me for a troll, which is 
too bad.

> I wouldn't say it's too much of a hassle calling IBM or ordering
> an AIX media kit online. Just one phonecall or just "a few clicks".

Why does it have to be ordered and paid for in the first place?

> Go buy a 44P or alike. They are more comparable to an old U-II
> machine in both price and performance.
> But then, why am I talking to a troll anyway?

Had you read my post carefully, the STIPULATION is that the machine be 
19" RACK MOUNTABLE. 44P is an old desktop system.

Please read more carefully next time, before you rush to label me as a 
troll. Thank you in advance.

Martin Etteldorf wrote:
> In comp.unix.aix UNIX admin <tripivceta@hotmail.com> wrote:
> 
>>Martin Etteldorf wrote:
>>
>>
>>>Well then... What exactly makes you thinking AIX wouldn't compete
>>>with Solaris? I know both since they first came out and I'd say that
>>>any of them has its own advantages and disadvantages.
>>>Please let us participate in your wisdom and tell us what makes Solaris
>>>so much better than AIX.
>>
>>Solaris is way more modern [...]
> 
> 
> Sounds interesting. Tell us more about this.
> 
> 
> 		Martin.
> 

Yes "UNIX Admin" - I too am dying to hear more.  If "Solaris is way more 
modern...", when (and how) did Solaris "obtain" the ability to create 
new LVs/FSs on the fly" (i.e., W/O having to bring the system down?) - 
umm, NATIVELY/"out of the box".

I too have been w/AIX almost from inception (ever hear of it running on 
a PS/2 - it did, I had to administer a few)...I wish I'd known "Solaris 
was WAY MORE MODERN" before Veritas came into a Sun admin's life!  Then 
again, why shell out the additional $$$$$$ to wannabe software vendors 
(like Veritas) when I get it for "free" (and THEN some) integrated w/AIX.

Yours...

Paul
p.s.  Ahhh - "hotmail.com" (shoulda noticed that to begin with)...PLONK

Martin Etteldorf wrote:

> Sounds interesting. Tell us more about this.

I have a better suggestion: instead of you baiting me, why don't you 
surf it up on http://store.sun.com/, http://docs.sun.com/ and 
http://sunsolve.sun.com/ and see for yourself?

Certainly if I'm a troll as you claim, I'm not to be trusted? So you'll 
have to do your homework for yourself, I guess.

The facts are there for everyone to read, download, and use.

Dohhhh <whynot@verizon.net> writes:
>Martin Etteldorf wrote:
>> In comp.unix.aix UNIX admin <tripivceta@hotmail.com> wrote:
>> 
>>>Martin Etteldorf wrote:
>>>
>>>
>>>>Well then... What exactly makes you thinking AIX wouldn't compete
>>>>with Solaris? I know both since they first came out and I'd say that
>>>>any of them has its own advantages and disadvantages.
>>>>Please let us participate in your wisdom and tell us what makes Solaris
>>>>so much better than AIX.
>>>
>>>Solaris is way more modern [...]
>> 
>> 
>> Sounds interesting. Tell us more about this.
>> 
>> 
>> 		Martin.
>> 
>
>Yes "UNIX Admin" - I too am dying to hear more.  If "Solaris is way more 
>modern...", when (and how) did Solaris "obtain" the ability to create 
>new LVs/FSs on the fly" (i.e., W/O having to bring the system down?) - 
>umm, NATIVELY/"out of the box".
>
>I too have been w/AIX 

AIX has the ODM, which suck starts Harley-Davidsons, and can therefore
safely be ignored.

-- 
       "The road to Paradise is through Intercourse."
        [email me at huge [at] huge [dot] org [dot] uk]

dirk dierickx wrote:
> ok, will that mean that starting from next week solaris will be a piece of
> crap? because you know, those same monkeys messing up linux could do the
> same to solaris? what are you going to use a year from now, when solaris
> is reduced to a pile of smelling amateur poo?

It is a possibility, yes. And I am worried about it, yes. Hope, however, 
lies in the fact that Sun will filter, regression test, debug, modify, 
then regression test again anything introduced from OpenSolaris back 
into Solaris.

So technically speaking, if I don't need to be at the bleeding edge, and 
I don't as long as I'm running the latest Solaris, I should be OK.

> anyway, you are avoiding my question! your view is that linux is badly
> written, mostly because, i don't know, you have this idea the people
> building it have no clue. i figure the same goes for the userland utils?

Correct. That is the thing that irks me the most when I have to work 
with Linux. It becomes very obvious right away that the guys who work on 
it are mostly clueless. And it's very bad because it makes the life of 
an engineer very hard.

> so if we take the OSS userland utils away from solaris, what will be left
> of your precious UNIX? would any unix still be the same without all the
> OSS tools?

But which OSS userland tools would you take away? Ones that have always 
been a part of Solaris, or the Freeware OSS userland tools? If we take 
the Freeware away, Solaris would still be a highly capable OS. But let 
me spin that around for you: if you take the OSS/Freeware out of Linux, 
what would be left? Could you do anything functional with that would be 
left? I could with Solaris. What about Linux?

> so, without further investigation, you decide that because they say it is
> cheaper, they have no clue what they are doing? perhaps for their
> individual case it _is_ the major advantage over others. in your thinking
> you would turn down a job at google because they use linux for the same
> reason (cheap).

You better believe I did a through investigation. I grilled these people 
back and forth, I took a tour of their facilities on several occasions, 
I looked at their systems, almost to the point where I interviewed them 
more than they interviewed me.

If they believe that Linux is cheap and are totally unaware of the fact 
that UNIX/Solaris is cheaper, then, yes, they are amateurs, and 
uninformed/misinformed amateurs to boot.

For $2000 I can get an UltraSPARC server with Solaris, and that is by no 
means more expensive than Linux. And that Solaris server will come with 
1yr support on SW and 3yrs support on HW. That's a deal.

Anyway, what can one expect from someone who would replace BIND "because 
tinyDNS is better". Or is it because nobody there really knew how to 
configure and maintain BIND properly? On the other hand, I've used 
tinyDNS, and it is very, very badly designed piece of SW.
So yes, these people are amateurs.

And as one good friend of mine puts it,

"I'll be a great team player if you put me on a team of skilled people, 
but I will NOT be a team player if you put me on a team with idiots!"

And the man is right.

> i'm sure they are proud, still the product contains bugs just like any
> other it is not perfect even though they are 'skilled profissional
> ENGINEERS'. the linux coders are proud as well, many of them work at hp,
> ibm, sgi, yes even sun, nasa, ... and their code is visible to _anybody_
> wishing to look at it.

It's not about bugs. It's abut conceptual design, which is an area where 
Linux is *extremely* bad. Even that would not be a problem, if Linux 
people actually had enough know-how to design it better. But they're 
completely and totally unaware of such concepts, and think they *know* 
what they're doing.
But just because it works, does not make it a quality product. That's 
called hacking.

Hacking is very, very bad in terms of structured engineering.

> but linux is not a unix, it is modeled after it, but it is something on
> its own.

I noticed.

dirk dierickx wrote:

> except SGI is doing it too with linux, i think their biggest machine sold
> was a 512 cpu 'box'.

You 'forgot' to mention that SGI does modifications that enable Linux to 
utilize the NUMA HW which allows for this. Luckily, I happen to know 
this since I used to work with Altix systems when they were still a 
prototype.

dirk dierickx wrote:

> sure, but the difference is that within oracle internally, they are
> dumping solaris in favour of linux.

Don't fool yourself. Historically, Oracle has steered its ship as the 
wind blew. If (Open)Solaris were to become a predominant platform 
because of x86/x64 in a few years' time, Larry would switch to it 
without thinking twice or batting an eyelash.

For Oracle, it's just pure business. Don't think Oracle is doing it 
because they genuinely believe in what Linux *tries* to stand for, or 
the Freeware/OSS principles. Oracle is not a revolutionary, but a 
corporation ruled by its CEO and a board of directors as well as 
shareholders and profits.

Dohhhh wrote:

> Yes "UNIX Admin" - I too am dying to hear more.  If "Solaris is way more 
> modern...", when (and how) did Solaris "obtain" the ability to create 
> new LVs/FSs on the fly" (i.e., W/O having to bring the system down?) - 
> umm, NATIVELY/"out of the box".

Solaris SVM, Sir. Please read the documentation before you get too 
worked up. It really is a good LVM.

Maybe it is I whould should PLONK you?

"UNIX admin" <tripivceta@hotmail.com> wrote in message 
news:42ac1944$0$1163$5402220f@news.sunrise.ch...
>
> This is OT, but before I abandon it as such, as I happen to work on cars 
> for years: BMW is not a top quality car. Piss on any car that you can't 
> even change oil on without having to reset the computer. That's no 
> quality, that's overpriced bullshit with a name tag on it.
>

You do are not forced to reset the computer after an oil change. The OBC is 
there for convenience. It monitors the number of engine starts, RPM, engine 
temp, miles driven and etc. Based on this it calculates when the oil needs 
to be changed and informs you as this time approaches. If you don't want to 
be warned... then fine, don't reset the computer. And it's not a "computer 
reset" as in a reboot or anything. You are simply telling the OBC that the 
oil has been changed so it can restart tracking this.

BMW is not the only make that tracks when the vehicle needs an oil change. 
If this is the biggest issue with a BMW has then it is certainly a quality 
car. At least this is what JD Powers and Associates believes:

#1 - Toyota
#2 - Honda
#3 - Porsche
#4 - BMW
#5 - Nissan

"UNIX admin" <tripivceta@hotmail.com> wrote in message 
news:42ac1a1c$0$1163$5402220f@news.sunrise.ch...
>
> I didn't look, I had to do it for a living. It put the bread on the table. 
> Working with Linux day to day was one of the worst jobs I ever held; I was 
> glad when I finally quit and went to work with Solaris again.
>

I work in what could be labeled as being a "Unix shop" but we also sell an 
NT version of our product because of customer demand. But for historical 
reasons our primary infrastructure runs on *nix. The sys-admin guys had a 
similar comment regarding Linux. They think it's great for specialized 
vertical applications but for doing general purpose stuff.... they don't 
like it at all. One of them mentioned that when it comes to system 
stability, Linux is worse than Solaris and (my flame retardant suit is on) 
Win2003 server.

I don't claim to administer servers so this is their opinion and not mine.

"UNIX admin" <tripivceta@hotmail.com> wrote in message 
news:42ac1fa6$0$1155$5402220f@news.sunrise.ch...
>
>> Can it run Doom 3 and UT 2004?
>
> It sure can, all you have to do is port those to Solaris.
>

To me this is so off the wall that it's almost OT.

Question: Where has Linux seen the most growth?
Answer: As a low cost server OS for mid-to-small organizations. The NIH 
(National Institute of Health) is an example. Will anyone be playing Doom or 
UT on that machine? I don't think so.

UNIX admin wrote:
> JEDIDIAH wrote:
> 
>>     Someone should tell Oracle that.
> 
> 
> Someone did tell Oracle that, and Oracle apparently listened, since
> Solaris10 x86/x64 is now officially supported.
> 
> BTW, Larry will certify any platform for his product if he believes will
> bring in more profit. There's nothing personal in his decisions.
> Having stated that, if Larry certified Solaris, it would imply that he
> deems it a profitable platform.
> 
> 
>>     They're dumping Solaris for it.
> 
> 
> See above.
> 
>>     Why run an OS that only scales to 100 cpus when I can run an OS
>> that scales to 1000?
> 
> 
> IRIX scales to thousands of CPUs, and so does Solaris. Linux does not,
> unless you use software clustering, which is not true CPU scaling.
> 

Do you know why most of the worlds super computers are using Linux?  Is
it because it's [perhaps] easier for desktop collaboration and running
of programs from various sources, most of whom use Linux on the desktop??

In comp.unix.aix UNIX admin <tripivceta@hotmail.com> wrote:
> Martin Etteldorf wrote:
> 
>> Sounds interesting. Tell us more about this.
> 
> I have a better suggestion: instead of you baiting me, why don't you 
> surf it up on http://store.sun.com/, http://docs.sun.com/ and 
> http://sunsolve.sun.com/ and see for yourself?

As already mentioned, I'm working with both AIX and solaris
for quite some years. Sorry, but I don't see where Solaris
should be the more modern OS.
I wouldn't say Solaris is bad. Quite the contrary. It's a
robust OS and it actually works. But then, that can be said
from AIX, too. Maybe I'm just too old for religious wars.

> [..] have to do your homework for yourself, I guess.

Been there, done that.


		Martin.
-- 
Remember to always speak softly.
But carry a big stick!

In comp.unix.aix UNIX admin <tripivceta@hotmail.com> wrote:
> Martin Etteldorf wrote:
>> Well, I'd say it's reasonably priced. A comparably powerful x86 or
>> sparc machine won't be cheaper.
> 
> I just built a 1U/19", 3.2GHz, 4GB RAM, 2x160GB, 2x1Gb Solaris10 server 
> based on P4/x86 for approx. $1800 USD or less.
> 
> If you can point me to a link that would sells a pSeries 1U/19" with the 
> latest AIX 5.x server for that amount of money, I will be paying CASH, 
> right away, NO QUESTIONS ASKED.

Maybe you should stay with your 32 bit machine instead if it is
sufficient for your needs.

>> Nonsense. Pay 35 bucks for the Media kit. Any RS/6000 comes with a
>> valid AIX license.
> 
> That's something. Could you post a link please?

Call your local IBM sales droid.

>> You're comparing a new pSeries machine with an obsolete Ultra-II
>> machine? Get a life.
> 
> Since when is B50 new? You really are taking me for a troll, which is 
> too bad.

A B50 for $3000? You're joking. They're selling for less than
500$ (or 400 Euros, for that matter) these days on ebay.


		Martin.

-- 
Remember to always speak softly.
But carry a big stick!

Larry Qualig wrote:

> I work in what could be labeled as being a "Unix shop" but we also sell an 
> NT version of our product because of customer demand. But for historical 
> reasons our primary infrastructure runs on *nix. The sys-admin guys had a 
> similar comment regarding Linux. They think it's great for specialized 
> vertical applications but for doing general purpose stuff.... they don't 
> like it at all. One of them mentioned that when it comes to system 
> stability, Linux is worse than Solaris and (my flame retardant suit is on) 
> Win2003 server.
> 
> I don't claim to administer servers so this is their opinion and not mine.

I'd just as soon pick Windows2003 server than Linux again.  It really is 
that bad. It's worse than that.

Larry Qualig wrote:
> To me this is so off the wall that it's almost OT.
> 
> Question: Where has Linux seen the most growth?
> Answer: As a low cost server OS for mid-to-small organizations. The NIH 
> (National Institute of Health) is an example. Will anyone be playing Doom or 
> UT on that machine? I don't think so.

The guy was trolling. But you can see the trend shifting: as soon as an 
OS hits a mainstream platform (x86/x64 at the moment), it will 
automatically start to be evaluated as a desktop OS. Solaris is no 
exception here, and neither would any other OS be.

KJ wrote:

> Do you know why most of the worlds super computers are using Linux?  Is
> it because it's [perhaps] easier for desktop collaboration and running
> of programs from various sources, most of whom use Linux on the desktop??

No it is not!
And yes I do know why.  It's because the scientific community has this 
misperception that Linux is CHEAP and AFFORDABLE, which is very, very wrong.

I say: let the scientists do the research, and engineers do the engineering.

Martin Etteldorf wrote:

> As already mentioned, I'm working with both AIX and solaris
> for quite some years. Sorry, but I don't see where Solaris
> should be the more modern OS.
> I wouldn't say Solaris is bad. Quite the contrary. It's a
> robust OS and it actually works. But then, that can be said
> from AIX, too. Maybe I'm just too old for religious wars.

I respect gray hairs, however, correct me if I wrong, but AIX runs only 
on IBM HW, true?

Whereas I just did a fresh install on a vanilla Taiwan server 
manufacturer and Solaris10 just picked every damn piece of HW in that 
machine up.  It just worked, as if it were running on Sun HW. And this 
was x86 HW.

Not only did Solaris10 just "magically" work, I found that, on the 
driver/doc CD shipped with the HW, there were OEM Solaris native 
packages of the drivers.
Not that I needed those, as Solaris already had them built in, but it's 
still nice to see that mainstream 3rd party OEMs are starting to support 
Solaris.

On Sun, 12 Jun 2005 12:32:33 +0200, UNIX admin
<tripivceta@hotmail.com> wrote in message
<<42ac0f2a$0$1147$5402220f@news.sunrise.ch>>:

> Of course they have, together with HP and SGI, and they will continue to 
> do so.
> 
> You see, these companies are betting that Linux will eventually grow in 
> popularity (over a span of several years) to the volumes of Windows. 

No, you don't see.  These companies are betting -- so far
successfully, at least in the case of HP and IBM -- that supporting
Linux will earn them a profit.

> You could also call it crawling on all fours and groveling.

No, that would be Sun's OpenSolaris and their "we're the openest of
open source and we have a Community Advisory Board who tells you so."

> Rather than firing their inept marketing goons and Microsoft prepaid 
> CEOs, they're choosing the path of the least resistence.

> Linux in its heart is a poorly hacked-together attempt of a UNIX like OS 
>   by people THAT DO NOT TRULY UNDERSTAND UNIX.

OpenSolaris IS NOT UNIX BUT YOU DON'T UNDERSTAND.  The OpenGroup
doesn't list Solaris 10 or any of its marketing derivates as
qualified.

> You can know every command there is to know in UNIX, but that does not 
> an engineer make!

A "UNIX admin" does not an engineer make.

On Sun, 12 Jun 2005 12:38:49 +0200, UNIX admin
<tripivceta@hotmail.com> wrote in message
<<42ac10a2$0$1147$5402220f@news.sunrise.ch>>:

> The only problem is, the price of the aging mainframes didn't drop 
> sufficiently, so IBM isn't seeing nearly the volumes sold that they'd 
> hoped to see.

Of course, you can prove this.

> It turns out IBM learned a hard lesson: not all that shines is gold. 
> Just because it's Linux doesn't make it an instant hit.

The lesson is actually that claiming falsely to support open source
won't save Sun.

On Sun, 12 Jun 2005 12:50:31 +0200, UNIX admin
<tripivceta@hotmail.com> wrote in message
<<42ac135f$0$1147$5402220f@news.sunrise.ch>>:

> No FUD please.

If FUD is disallowed, you'll have to stop posting altogether.
That would be a good thing.

On Sun, 12 Jun 2005 13:04:23 +0200, UNIX admin
<tripivceta@hotmail.com> wrote in message
<<42ac16a0$0$1163$5402220f@news.sunrise.ch>>:

> IRIX scales to thousands of CPUs, and so does Solaris. Linux does not, 
> unless you use software clustering, which is not true CPU scaling.

The ADMIN proclaims itself an engineer.

On Sun, 12 Jun 2005 13:09:39 +0200, UNIX admin
<tripivceta@hotmail.com> wrote in message
<<42ac17dc$0$1163$5402220f@news.sunrise.ch>>:

> If you are implying that Linux is a fine operating system, I do emplore 
> you to delve deeper into operating systems, so that one day you may be 
> able to tell the difference.

You've demonstrated you lack of knowledge of operating system design.

> Knowledge is power, as they say.

You're completely powerless.

On Sun, 12 Jun 2005 13:15:39 +0200, UNIX admin
<tripivceta@hotmail.com> wrote in message
<<42ac1944$0$1163$5402220f@news.sunrise.ch>>:

> This is OT, but before I abandon it as such, as I happen to work on cars 
> for years:

So you're only a UNIX admin after your job as a mechanic.

On Sun, 12 Jun 2005 13:22:09 +0200, UNIX admin
<tripivceta@hotmail.com> wrote in message
<<42ac1ac9$0$1163$5402220f@news.sunrise.ch>>:

> It's quite simple, actually: OpenSolaris will be like RedHat's Fedora 
> core in terms of product placement.

In terms of product placement, Sun's market share will still be
miniscule.

On Sun, 12 Jun 2005 13:27:35 +0200, UNIX admin
<tripivceta@hotmail.com> wrote in message
<<42ac1c10$0$1156$5402220f@news.sunrise.ch>>:

> Liam Slider wrote:
> 
> > Novell isn't "name brand, officially certified"? Redhat isn't?
> 
> Not in terms of UNIX.

In terms of UNIX, OpenSolaris isn't certified.

> As it stands right now, no Linux distro/kernel is 
> able to be UNIX certified.

As it stands right now, OpenSolaris isn't certified.

> The ironic thing is

The ironic thing is you're an idiot repeating marketing drivel while
purporting to have technical knowledge.

On Sun, 12 Jun 2005, UNIX admin wrote:

> Not that I needed those, as Solaris already had them built in, but it's
> still nice to see that mainstream 3rd party OEMs are starting to support
> Solaris.

It is.  Although there's still a lot of work to be done, the day's
of SOlaris x86's pitiful HW support are now behind us.

FWIW, I'm typing this on a Acer Ferrari 3400 x64 laptop.  Running
Solaris 10, of course.

-- 
Rich Teer, SCNA, SCSA, OpenSolaris CAB member

President,
Rite Online Inc.

Voice: +1 (250) 979-1638
URL: http://www.rite-group.com/rich

On Sun, 12 Jun 2005, Ku Karlovsky wrote:

> The lesson is actually that claiming falsely to support open source
> won't save Sun.

Why do you people continue to claim this?  SUn has given more to
open source than any other entity, save UCB.  ANd that will change
with the release of Open SOlaris.

-- 
Rich Teer, SCNA, SCSA, OpenSolaris CAB member

President,
Rite Online Inc.

Voice: +1 (250) 979-1638
URL: http://www.rite-group.com/rich

Rich Teer wrote:
> On Sun, 12 Jun 2005, UNIX admin wrote:
> 
> 
>>Not that I needed those, as Solaris already had them built in, but it's
>>still nice to see that mainstream 3rd party OEMs are starting to support
>>Solaris.
> 
> 
> It is.  Although there's still a lot of work to be done, the day's
> of SOlaris x86's pitiful HW support are now behind us.
> 
> FWIW, I'm typing this on a Acer Ferrari 3400 x64 laptop.  Running
> Solaris 10, of course.
> 

I hate you. Where do I get one of those? (laptops that is. I love my 
Ferraris).

-- 
Cheers,					http://www.dotware.co.uk
Jim			  http://www.dotware-entertainment.co.uk

The mind of a bigot is like the pupil of the eye. The more light
you shine on it, the more it will contract.
     - Oliver Wendell Holmes Jr.

On Sun, 12 Jun 2005 17:47:16 +0200, UNIX admin wrote:

> No it is not!
> And yes I do know why.  It's because the scientific community has this 
> misperception that Linux is CHEAP and AFFORDABLE, which is very, very wrong.
> 
> I say: let the scientists do the research, and engineers do the engineering.

If Linux is so much *worse* then Unix, and the issue is only some
appairently presumed lower cost...then why is it that not only do *most*
supercomputers run Linux, but all the *fastest* ones as well? 

Here in comp.unix.solaris,
UNIX admin <tripivceta@hotmail.com> spake unto us, saying:

>It's not about bugs. It's abut conceptual design, which is an area where 
>Linux is *extremely* bad.

While I realize this isn't an advocacy group, I'm curious if you could
clarify the above comment.

I see such things spouted from the *BSD camp from time to time, but a
number of their opinions seem strongly rooted in license philosophy
differences and not in the actual characteristics of the software.

Are you referring to the Linux kernel itself, or to other elements of a
typical Linux system?  And in which context?

>But just because it works, does not make it a quality product. That's 
>called hacking.

While true, a highly-centralized and structured process by itself does
not guarantee that a product will be useful, and it can also limit the
ability of a platform to adjust to changing user requirements.

It's a rare platform that is both structured and flexible, and I'd love
to find such a beast.  :-)

-- 
 -Rich Steiner >>>---> http://www.visi.com/~rsteiner >>>---> Smyrna, GA USA
  OS/2 + eCS + Linux + Win95 + DOS + PC/GEOS + Executor = PC Hobbyist Heaven!
       WARNING: I've seen FIELDATA FORTRAN V and I know how to use it!
                   The Theorem Theorem: If If, Then Then.

UNIX admin wrote:
> KJ wrote:
> 
>> Do you know why most of the worlds super computers are using Linux?  Is
>> it because it's [perhaps] easier for desktop collaboration and running
>> of programs from various sources, most of whom use Linux on the desktop??
> 
> 
> No it is not!
> And yes I do know why.  It's because the scientific community has this
> misperception that Linux is CHEAP and AFFORDABLE, which is very, very
> wrong.
> 
> I say: let the scientists do the research, and engineers do the
> engineering.
> 
After I sent that question, I started to wonder if if was because IBM is
building so many of them these days?

In article <oCVqe.2273$ej1.2095@trndny09>,
	Dohhhh <whynot@verizon.net> writes:
> Yes "UNIX Admin" - I too am dying to hear more.  If "Solaris is way more 
> modern...", when (and how) did Solaris "obtain" the ability to create 
> new LVs/FSs on the fly" (i.e., W/O having to bring the system down?) - 
> umm, NATIVELY/"out of the box".

LV's: Solaris 2.5, about 10 years ago IIRC.
FS's: I'm pretty sure this was in the first Sun OS I ever used,
      SunOS 2 about 20 years ago.

-- 
Andrew Gabriel

In comp.unix.aix UNIX admin <tripivceta@hotmail.com> wrote:
> Martin Etteldorf wrote:
> 
>> As already mentioned, I'm working with both AIX and solaris
>> for quite some years. Sorry, but I don't see where Solaris
>> should be the more modern OS.
>> I wouldn't say Solaris is bad. Quite the contrary. It's a
>> robust OS and it actually works. But then, that can be said
>> from AIX, too. Maybe I'm just too old for religious wars.
> 
> I respect gray hairs, however, correct me if I wrong, but AIX runs only 
> on IBM HW, true?

False. It runs on PPC and POWER only, but not limited to IBM.

> Whereas I just did a fresh install on a vanilla Taiwan server 
> manufacturer and Solaris10 just picked every damn piece of HW in that 
> machine up.  It just worked, as if it were running on Sun HW. And this 
> was x86 HW.

Try installing it on a Board with an ATI RS300/IXP200 Chipset.


		Martin.
-- 
Remember to always speak softly.
But carry a big stick!

On Sunday 12 June 2005 06:11, UNIX admin <tripivceta@hotmail.com>
(<42ac2644$0$1149$5402220f@news.sunrise.ch>) wrote:

> I have a better suggestion: instead of you baiting me, why don't you
> surf it up on http://store.sun.com/, http://docs.sun.com/ and
> http://sunsolve.sun.com/ and see for yourself?

I have an even better suggestion: instead of cross-posting your marketing
poop, stick to Solaris newsgroups where it might be on-topic.

> Certainly if I'm a troll as you claim, I'm not to be trusted?

You are certainly a troll and certainly not to be trusted.

> So you'll have to do your homework for yourself, I guess.

Someone, like you, who falls for Mike Cox trolls, is quite gullible.  You'll
need to learn to do your own homework for yourself, I know.

In article <d8i39f$476$1@schlund.de>,
Martin Etteldorf  <etteldor@mail.ru> wrote:
>>> I wouldn't say Solaris is bad. Quite the contrary. It's a
>>> robust OS and it actually works. But then, that can be said
>>> from AIX, too. Maybe I'm just too old for religious wars.
>> 
>> I respect gray hairs, however, correct me if I wrong, but AIX runs only 
>> on IBM HW, true?
>
>False. It runs on PPC and POWER only, but not limited to IBM.

Show me AIX running on a PowerMAC, show me AIX running on a Genesi Pegasos.

-- 
EMail:joerg@schily.isdn.cs.tu-berlin.de (home) J�rg Schilling D-13353 Berlin
      js@cs.tu-berlin.de		(uni)  
      schilling@fokus.fraunhofer.de	(work) Blog: http://schily.blogspot.com/
URL:  http://cdrecord.berlios.de/old/private/ ftp://ftp.berlios.de/pub/schily

Ku Karlovsky wrote:

> The ironic thing is you're an idiot repeating marketing drivel while
> purporting to have technical knowledge.

The ironic thing is, you have no technical merit.

Ku Karlovsky wrote:

> You've demonstrated you lack of knowledge of operating system design.

How exactly have I demonstrated that?

> You're completely powerless.

Which is why you're trying to provoke me?  OK.

Ku Karlovsky wrote:

> In terms of product placement, Sun's market share will still be
> miniscule.

A true SVR4/POSIX UNIX is coming, and it is free as in beer.

It is over for you Linux folks.  It is pointless to resist; give it up.

begin  virus.txt.scr Liam Slider wrote:

> On Sun, 12 Jun 2005 17:47:16 +0200, UNIX admin wrote:
> 
>> No it is not!
>> And yes I do know why.  It's because the scientific community has this
>> misperception that Linux is CHEAP and AFFORDABLE, which is very, very
>> wrong.
>> 
>> I say: let the scientists do the research, and engineers do the
>> engineering.
> 
> If Linux is so much *worse* then Unix, and the issue is only some
> appairently presumed lower cost...then why is it that not only do *most*
> supercomputers run Linux, but all the *fastest* ones as well?

Sour grapes syndrome?
-- 
There are two kinds of people in this world: the kind that divides
everybody  into two kinds of people, and everybody else

rsteiner@visi.com (Richard Steiner) writes:

> Here in comp.unix.solaris,
> UNIX admin <tripivceta@hotmail.com> spake unto us, saying:
>
>>It's not about bugs. It's abut conceptual design, which is an area where 
>>Linux is *extremely* bad.
>
> While I realize this isn't an advocacy group, I'm curious if you could
> clarify the above comment.
>

Threads? It took Linus years to realise that the LinuxThreads design, based
on clone() is not good enough, even though they claimed it's superior to
the traditional thread design. They finally switched to NTPL some time
ago. If you monitor Linux developers and/or threading newgroups, you will
notice that there are now a lot of posting of the type "My application,
which did work under LinuxThreads doesn't work any more".

Interestingly enough, these application writers didn't care that their
application didn't run under any other POSIX threads capable OS claiming
they write for Linux only anyway and don't care about the others.

Now they have to redesign it as it doesn't run under Linux either. Revenge
is sweet :-)


Dragan

-- 
Dragan Cvetkovic, 

To be or not to be is true. G. Boole      No it isn't.  L. E. J. Brouwer

!!! Sender/From address is bogus. Use reply-to one !!!

In article <d8i5df$9jr$1@news.cs.tu-berlin.de>, Joerg Schilling <js@cs.tu-berlin.de> wrote:
> In article <d8i39f$476$1@schlund.de>,
> Martin Etteldorf  <etteldor@mail.ru> wrote:
>>>> I wouldn't say Solaris is bad. Quite the contrary. It's a
>>>> robust OS and it actually works. But then, that can be said
>>>> from AIX, too. Maybe I'm just too old for religious wars.
>>> 
>>> I respect gray hairs, however, correct me if I wrong, but AIX runs only 
>>> on IBM HW, true?
>>
>>False. It runs on PPC and POWER only, but not limited to IBM.
>
> Show me AIX running on a PowerMAC, show me AIX running on a Genesi Pegasos.

I should point out that AIX has run on commercially sold systems such as
Bull (Escala line: http://www.bull.com/aix/index.html) and Motorola
(PowerStack line but now discontinued).

-Dan

In article <slrndape8e.8fj.usenet@gaia.roc2.gblx.net>, Dan Foster <usenet@evilphb.org> wrote:
> In article <d8i5df$9jr$1@news.cs.tu-berlin.de>, Joerg Schilling <js@cs.tu-berlin.de> wrote:
>> In article <d8i39f$476$1@schlund.de>,
>> Martin Etteldorf  <etteldor@mail.ru> wrote:
>>>>> I wouldn't say Solaris is bad. Quite the contrary. It's a
>>>>> robust OS and it actually works. But then, that can be said
>>>>> from AIX, too. Maybe I'm just too old for religious wars.
>>>> 
>>>> I respect gray hairs, however, correct me if I wrong, but AIX runs only 
>>>> on IBM HW, true?
>>>
>>>False. It runs on PPC and POWER only, but not limited to IBM.
>>
>> Show me AIX running on a PowerMAC, show me AIX running on a Genesi Pegasos.
>
> I should point out that AIX has run on commercially sold systems such as
> Bull (Escala line: http://www.bull.com/aix/index.html) and Motorola
> (PowerStack line but now discontinued).

I forgot to add that AIX 4.1.4 (I think) also ran on Apple Power
Macintosh servers under the 'Apple Network Server' name. Apple
discontinued this product years ago.

Earliest releases of AIX also ran on the IBM PS/2 (even though this is
not POWER/PowerPC-based) but was discontinued prior to AIX version 3.

AIX had also been ported to Intel's Itanium ("Project Monterey" with
SCO) but that never made it to the marketplace.

-Dan

UNIX admin wrote:
> Kevin Bowling wrote:
> 
>> They have a bunch of them on the zSeries, yet still ported Linux to it..
> 
> 
> That's because IBM assumed (and correctly so) that a mainstream OS like
> Linux would make their aging mainframes more marketable.
> 
> The only problem is, the price of the aging mainframes didn't drop
> sufficiently, so IBM isn't seeing nearly the volumes sold that they'd
> hoped to see.
  Aging?  These are built out of the very CPUs that made HP send Alpha
and HPPA to the graveyard, SGI to all but drop MIPS and Sun look like a
fool with their Sparc promises (POWER).  Point in case, big iron of this
type isn't for everyone and the people who use it didn't need yet
another OS (IE VISA isn't running Linux on zSeries..).  It is mainly
used as a connector for webapps to access zSeries information as the
other OSes are more "competitive".  I highly doubt IBM gained anything
by porting Linux to zSeries, nor was it their primary motivation.

On Sun, 12 Jun 2005, Jim wrote:

> I hate you. Where do I get one of those? (laptops that is. I love my
> Ferraris).

I got mine from my local ACer dealer.  It's not something they keep in
stock, so they had to order one for me.  It took about a week to
arrive.

-- 
Rich Teer, SCNA, SCSA, OpenSolaris CAB member

President,
Rite Online Inc.

Voice: +1 (250) 979-1638
URL: http://www.rite-group.com/rich

UNIX admin wrote:
> 
> Meanwhile, Ultra Enterprise450, 220R, 250R, and 280Rs are selling for
> $400-$600 USD on ebay, and I can get Solaris just with a few clicks,
> hassle free.
You are comparing apples to oranges in Price/Performance ratios here.
In addition, lack of newer (and older, even MCA) RS/6000s on eBay is
beacuse they are still doing their JOB, not needing to be replaced every
couple years like other garbage.

On Sunday 12 June 2005 14:19, Joerg Schilling <js@cs.tu-berlin.de>
(<d8i5df$9jr$1@news.cs.tu-berlin.de>) wrote:

> Show me AIX running on a PowerMAC, show me AIX running on a Genesi
> Pegasos.

Show me OpenSolaris with Unix 2003 certification.

UNIX admin wrote:
> Arkady Duntov wrote:

>> Apparently OpenSolaris isn't Unix either.

> Since the first release of OpenSolaris will be a practically 1:1 copy of 
> Solaris proper, and Solaris is UNIX, it therefore follows that 
> OpenSolaris is UNIX.
> 
> Just to make sure you don't play dumb any more:
> 
> if A = B and B = C it follows that A = C, or,
> if A = C and A = B then B = C.

This depends on what definition of Unix you are talking about.  If
you are talking about The-Open-Group-trademark-worthiness, then
OpenSolaris probably isn't Unix.  If you defined the property D(B,A)
to mean "B is derived from A", and the properly U(A) to mean "A is
certified as UNIX(r) by The Open Group", then U(A) & D(B,A) does not
imply U(B).

On the other hand, if you wanted to be a little more pragmatic,
you could define the property E(A) to mean "A, if it were tested,
would be easily pass or would nearly pass the test for certification
by The Open Group", then you could say U(A) & D(B,A) implies E(B).

So, the question then is, what does "is Unix" mean?  Does it mean D(B)
or does it mean E(B)?  Or does it mean "looks like Unix, smells like
Unix, and tastes like Unix"?  That's more of a philosophical question,
so we probably won't get anywhere debating it.

   - Logan

In comp.unix.solaris Kamal R. Prasad <kamalp@acm.org> wrote:

> Just to clear some misunderstandings (not on behalf of IBM as i don't
> work for them) -AIX was NOT developed by IBM. It was developed by Bull
> SA of France.

> regards
> -kamal


Source ? When, then, did IBM acquire AIX from Bull and take
over the development ?

I just find it hard to believe, as 
it was quite the open secret that Bull used IBM's RISC/6000
workstations (rebadged) and IBM delivered a custom version
of AIX to Bull that had all references to "AIX" and "IBM"
replaced with "BullOS" and "Bull". Certainly there was a
great deal of collaboration on the part of Bull with IBM
to satisfy Bull's custom product needs, but it was IBM that
led the AIX development.

Of course this was over 10 years ago.

Kevin Bowling wrote:
> It is mainly
> used as a connector for webapps to access zSeries information as the
> other OSes are more "competitive".  I highly doubt IBM gained anything
> by porting Linux to zSeries, nor was it their primary motivation.

Exactly, if i remember correctly the whole Linux on zSeries thing
was in fact something like a mutiny done by the folks in B�blingen/
Germany. They kept the portation a secret until it was finished,
supposedly because it would otherwise have been stopped by "upper
managment".

Regards,

	Frank

UNIX admin wrote:
> The guy was trolling. But you can see the trend shifting: as soon as an
> OS hits a mainstream platform (x86/x64 at the moment), it will
> automatically start to be evaluated as a desktop OS. Solaris is no
> exception here, and neither would any other OS be.

  He was trolling?  YOU are trolling, just look at all your incumbent
posts agitating arguments based on Sun marketing talking points.  No
shit it will run if you port it..

In comp.unix.aix Joerg Schilling <js@cs.tu-berlin.de> wrote:
> In article <d8i39f$476$1@schlund.de>,
> Martin Etteldorf  <etteldor@mail.ru> wrote:
>>>> I wouldn't say Solaris is bad. Quite the contrary. It's a
>>>> robust OS and it actually works. But then, that can be said
>>>> from AIX, too. Maybe I'm just too old for religious wars.
>>> 
>>> I respect gray hairs, however, correct me if I wrong, but AIX runs only 
>>> on IBM HW, true?
>>
>>False. It runs on PPC and POWER only, but not limited to IBM.
> 
> Show me AIX running on a PowerMAC,

Apple Network Server 500/700

> show me AIX running on a Genesi Pegasos.

I didn't write it runs on _any_ PPC hardware, did I?
That said, show me Solaris running on a IBM NetVista 2200, which is
obviously a x86 based system.



		Martin.

-- 
Remember to always speak softly.
But carry a big stick!

On Sun, 12 Jun 2005 12:41:25 +0200, UNIX admin
<tripivceta@hotmail.com> wrote:

>Martin Etteldorf wrote:
>
>> Well then... What exactly makes you thinking AIX wouldn't compete
>> with Solaris? I know both since they first came out and I'd say that
>> any of them has its own advantages and disadvantages.
>> Please let us participate in your wisdom and tell us what makes Solaris
>> so much better than AIX.
>
>Solaris is way more modern and supports way more hardware. Unlike AIX, 
>it kept up with the times and it also sprung to two new platforms 
>besides UltraSPARC/UPA/PCI, namely x86 and x64.
I love Solaris, but Aix is also a excelent operatings system, i
administer a Aix box a few years ago.

We want to buy a new UltraSparc machine, but the management wanted a
cheaper alternative for the UltraSparc, there wanted Windows servers,
but the IT department made clear, that is crazy to do on a Windows
server, what we do on Solaris, we running a lot of programs on one
server. But a Intel or a  AMD server is appealing, the performance
from the Ultra Sparc is far behind the cpu's  from Intel, AMD and
PowerPC. Sun lost the cpu war, on this moment. The new Operon servers
from Sun are attractive , but most of our Software is not running and
supported on Solaris X86,  the only alternative we have is to put
Linux enterprise edition from Redhat or  Suse on this machine, our
software is supported on Linux.  A pity, i find Solaris, a beter
operatingssystem. But this matter is not decided yet,  the IT
deparment is favor for the Sparc.

In article <d8j1d6$ua$1@home.itg.ti.com>,
	Chris Barrera <cbarrera@dontspamme.com> writes:
> In comp.unix.solaris Kamal R. Prasad <kamalp@acm.org> wrote:
> 
>> Just to clear some misunderstandings (not on behalf of IBM as i don't
>> work for them) -AIX was NOT developed by IBM. It was developed by Bull
>> SA of France.
> 
> Source ? When, then, did IBM acquire AIX from Bull and take
> over the development ?

I don't know what involvement Bull had, but AIX was first developed
by Interactive Systems Corporation under contract from IBM, as the
reference port of SVR3.1 (IIRC) to one of IBM's systems (I guess the
RS6000, but I don't know that for sure). These were the same folks who
produced Interactive UNIX as their own product, and did SVR3 ports
to other platforms under contract from their respective manufacturers.

After a couple of aquisitions, Sun ended up with Interactive UNIX and
lots of the former ISC staff, many of whom are still at Sun. They
were still in the old ISC offices in Los Angeles until about a year
ago when Sun finally moved them to another Sun office down the road.

-- 
Andrew Gabriel

Chris Barrera wrote:
> In comp.unix.solaris Kamal R. Prasad <kamalp@acm.org> wrote:
>
> > Just to clear some misunderstandings (not on behalf of IBM as i don't
> > work for them) -AIX was NOT developed by IBM. It was developed by Bull
> > SA of France.
>
> > regards
> > -kamal
>
>
> Source ? When, then, did IBM acquire AIX from Bull and take
> over the development ?
>
The best publicly available source I can think of is:-
http://febcm.club.fr/english/chronoa13.htm

Other than that, I happened to work at IBM India and have seen src code
with a multi-year copyright + license to use from Bull to IBM. I can
conlude safely that quite a bit if not all of AIX src code was
developed by Bull and is still being maintained under contract by some
appendage of Bull.

> I just find it hard to believe, as

I find it easier to believe that IBM would depend on someone else to
produce AIX (about the same way they depended on MSFT to develop an OS
for a desktop computer).

> it was quite the open secret that Bull used IBM's RISC/6000
> workstations (rebadged) and IBM delivered a custom version
> of AIX to Bull that had all references to "AIX" and "IBM"
> replaced with "BullOS" and "Bull". Certainly there was a

Probably not. I mean, they must have marketed it as AIX itself and Bull
was just a vendor of computers, with IBM as a chip supplier.

> great deal of collaboration on the part of Bull with IBM
> to satisfy Bull's custom product needs, but it was IBM that
> led the AIX development.
>
Im not sure I understand what customization Bull had to go through.
They weren't into retrofitting computers to defence equipment or
something like that. Their target customer base was the same that buys
unix systems today.

regards
-kamal

Liam Slider wrote:
> On Sun, 12 Jun 2005 17:47:16 +0200, UNIX admin wrote:
> 
> 
>>No it is not!
>>And yes I do know why.  It's because the scientific community has this 
>>misperception that Linux is CHEAP and AFFORDABLE, which is very, very wrong.
>>
>>I say: let the scientists do the research, and engineers do the engineering.
> 
> 
> If Linux is so much *worse* then Unix, and the issue is only some
> appairently presumed lower cost...then why is it that not only do *most*
> supercomputers run Linux, but all the *fastest* ones as well? 
> 

The same reason why "everyone" is using Microsoft Windows on their pc's?

Trond

tripivceta@hotmail.com wrote:

>> Is this is the unix distro with CDDL crap bordering fake
>> open source license that all of 1350 people downloaded?
> 
> No, it's a true SVR4, POSIX-compliant, backward and forward compatible,
> enterprise-class UNIX being open sourced for the first time in history.

"open sourced?" People have been licensing the source to UNIX for a long
time.
> 
> Up until now we had amateur crap like Linux released for free, there
> was priactuically no choice, but now there'll be a true enterprise
> class, PROFESSIONAL UNIX available for free.

It's funny, if you look at any UNIX system, MOST of it will be open source
and free code.

> 
> Hmmm, which one should I pick:
> 
> Linux, an amateur OS written by amateurs for amateurs

Linux, written by engineers for engineers.

> 
> or
> 
> (Open)Solaris, a professionally ENGINEERED operating ENVIRONMENT?

Designed to meet marketing memos and deadlines, leaving buzz-word complient
pseudo-features in, but cutting real development as being risky.

> 
> Gee, what a dilemma.

Yes.

On Mon, 13 Jun 2005 16:03:45 +0200, Trond Norbye wrote:

> The same reason why "everyone" is using Microsoft Windows on their pc's?

No, I don't think Linux comes pre-installed on supercomputers....


Nor is some giant company using arm twisting tactics to ensure it's put on
supercomputers.


So you'd be wrong.

>> The same reason why "everyone" is using Microsoft Windows on their pc's?

> No, I don't think Linux comes pre-installed on supercomputers....


That's funny - and with just the right amount of sarcasm. This would
make for a great Far Side or Dilbert cartoon.

Kevin Bowling wrote:

>   He was trolling?  YOU are trolling, just look at all your incumbent
> posts agitating arguments based on Sun marketing talking points.

Oh, it's agitating you, is it?  Were a Solaris user, then it wouldn't.

> No shit it will run if you port it..

So if you are in agreement with me, then we have no problem.

Larry Qualig wrote:
> "UNIX admin" <tripivceta@hotmail.com> wrote in message
> news:42ac1fa6$0$1155$5402220f@news.sunrise.ch...
> >
> >> Can it run Doom 3 and UT 2004?
> >
> > It sure can, all you have to do is port those to Solaris.
> >
>
> To me this is so off the wall that it's almost OT.
>
> Question: Where has Linux seen the most growth?
> Answer: As a low cost server OS for mid-to-small organizations. The NIH
> (National Institute of Health) is an example. Will anyone be playing Doom or
> UT on that machine? I don't think so.

I use linux as my desktop OS. It also runs several databases and
application servers, but still a desktop OS.

UNIX admin wrote:
> > Does OpenSolaris' X server run faster? Does it support composite?
>
> It runs X.org, so it at least runs at the same speed or faster.

Then it will have the same problems.

On Mon, 13 Jun 2005 18:43:54 +0200, UNIX admin
<tripivceta@hotmail.com> wrote in message
<<42adb7af$0$1159$5402220f@news.sunrise.ch>>:

> Oh, it's agitating you, is it?  Were a Solaris user, then it wouldn't.

"UNIX admin" is the Oxtard of Solaris.

On Mon, 13 Jun 2005 09:01:49 -0700, lqualig wrote:

>>> The same reason why "everyone" is using Microsoft Windows on their pc's?
> 
>> No, I don't think Linux comes pre-installed on supercomputers....
> 
> 
> That's funny - and with just the right amount of sarcasm. This would
> make for a great Far Side or Dilbert cartoon.

I was thinking more along the lines of User Friendly myself...

On 2005-06-12, Andrew Gabriel <andrew@cucumber.demon.co.uk> wrote:
> In article <oCVqe.2273$ej1.2095@trndny09>,
> 	Dohhhh <whynot@verizon.net> writes:
>> Yes "UNIX Admin" - I too am dying to hear more.  If "Solaris is way more 
>> modern...", when (and how) did Solaris "obtain" the ability to create 
>> new LVs/FSs on the fly" (i.e., W/O having to bring the system down?) - 
>> umm, NATIVELY/"out of the box".
>
> LV's: Solaris 2.5, about 10 years ago IIRC.

	Solaris 2.5 or some release of Veritas that supported Solaris 2.5?

> FS's: I'm pretty sure this was in the first Sun OS I ever used,
>       SunOS 2 about 20 years ago.
>


-- 
	The best OS in the world is ultimately useless         |||
	if it is controlled by a Tramiel, Jobs or Gates.      / | \
                                                     

On 2005-06-12, UNIX admin <tripivceta@hotmail.com> wrote:
> Martin Etteldorf wrote:
>
>> As already mentioned, I'm working with both AIX and solaris
>> for quite some years. Sorry, but I don't see where Solaris
>> should be the more modern OS.
>> I wouldn't say Solaris is bad. Quite the contrary. It's a
>> robust OS and it actually works. But then, that can be said
>> from AIX, too. Maybe I'm just too old for religious wars.
>
> I respect gray hairs, however, correct me if I wrong, but AIX runs only 
> on IBM HW, true?
>
> Whereas I just did a fresh install on a vanilla Taiwan server 
> manufacturer and Solaris10 just picked every damn piece of HW in that 
> machine up.  It just worked, as if it were running on Sun HW. And this 
> was x86 HW.

	Been there did that. Never want to go there again ever.

[deletia]

	If you're not running Solaris on a vertically integrated piece of
hardware with a planetary support apparatus behind it you might as well be
using the version of SVR4 written by Atari to run on the TT.

	A "clone" takes the whole point of putting up with Sun and
disposes of it.

-- 
	The best OS in the world is ultimately useless         |||
	if it is controlled by a Tramiel, Jobs or Gates.      / | \
                                                     

On 2005-06-12, UNIX admin <tripivceta@hotmail.com> wrote:
> JEDIDIAH wrote:
>
>> 	AIX is quite competitive. It also doesn't require quite so
>> many extra cost addons as Solaris does in order to be fully 
>> functional.
>
> No FUD please. Solaris is fully functional out-of-the-box, and even 
> advanced tools like SVM come for free with the OS out-of-the-box.
>

	MY Solaris admins seem to think differently.

	...nothing like working on AIX to take the fan out of Sun fan-boy.

-- 
	The best OS in the world is ultimately useless         |||
	if it is controlled by a Tramiel, Jobs or Gates.      / | \
                                                     

On 2005-06-12, UNIX admin <tripivceta@hotmail.com> wrote:
> JEDIDIAH wrote:
>> 	Someone should tell Oracle that.
>
> Someone did tell Oracle that, and Oracle apparently listened, since 
> Solaris10 x86/x64 is now officially supported.
>
	
	...been "officially supported" in the past. Many of us still
remember those days.


> BTW, Larry will certify any platform for his product if he believes will 
> bring in more profit. There's nothing personal in his decisions.

	Larry doesn't merely "certify" Linux. It is his flagship high 
performance platform.
	

> Having stated that, if Larry certified Solaris, it would imply that he 
> deems it a profitable platform.
>
>
>> 	They're dumping Solaris for it.
>
> See above.

	What you said above doesn't have any relevance.

	Linux is Larry's favorite son now. No amount of spin control will
change that.

>
>> 	Why run an OS that only scales to 100 cpus when I can run an OS
>> that scales to 1000?
>
> IRIX scales to thousands of CPUs, and so does Solaris. Linux does not, 
> unless you use software clustering, which is not true CPU scaling.

	This is a bald faced lie on both counts?

	Have you ever even touched a fireplane system? I doubt it.

-- 
	The best OS in the world is ultimately useless         |||
	if it is controlled by a Tramiel, Jobs or Gates.      / | \
                                                     

On 2005-06-13, Trond Norbye <trond.norbye@gmail.com> wrote:
> Liam Slider wrote:
>> On Sun, 12 Jun 2005 17:47:16 +0200, UNIX admin wrote:
>> 
>> 
>>>No it is not!
>>>And yes I do know why.  It's because the scientific community has this 
>>>misperception that Linux is CHEAP and AFFORDABLE, which is very, very wrong.
>>>
>>>I say: let the scientists do the research, and engineers do the engineering.
>> 
>> 
>> If Linux is so much *worse* then Unix, and the issue is only some
>> appairently presumed lower cost...then why is it that not only do *most*
>> supercomputers run Linux, but all the *fastest* ones as well? 
>> 
>
> The same reason why "everyone" is using Microsoft Windows on their pc's?

	So you would rather just slander every scientist on the planet.

-- 
	The best OS in the world is ultimately useless         |||
	if it is controlled by a Tramiel, Jobs or Gates.      / | \
                                                     

On 2005-06-13, lqualig@uku.co.uk <lqualig@uku.co.uk> wrote:
>>> The same reason why "everyone" is using Microsoft Windows on their pc's?
>
>> No, I don't think Linux comes pre-installed on supercomputers....
>
>
> That's funny - and with just the right amount of sarcasm. This would
> make for a great Far Side or Dilbert cartoon.
>

	You can get prebuilt computing clusters with Linux intalled actually...

-- 
	The best OS in the world is ultimately useless         |||
	if it is controlled by a Tramiel, Jobs or Gates.      / | \
                                                     

On 2005-06-12, UNIX admin <tripivceta@hotmail.com> wrote:
> JEDIDIAH wrote:
>
>> 	4- historically bad x86 support.	
>
> Not any more. Next!

	Changing history. How Orwellian of you.

>
>> 	5- 3rd party vendors abandoning it in favor of Linux.
>
> Not any more either: RealTek supports it out of the box, so does intel, 
> so does nVidia, etcetera. Next!
>
>> 	6- Incomplete compared to other RISC Unixen.
>
> What exactly do you find incomplete?
>
>> 	7- Scales to fewer cpus than "hobbyist" competitors.
>
> Solaris easily runs on 1024 CPUs WITHOUT CLUSTERING. Next!

	That would certainly be news to the rest of us.

	The SGI Altix will scale to 1024 cpus running Linux and a single
NUMA image of the operating system.

	What Sun/Fujistu hardware does this for Solaris?

>
>> 	8- runs on slowest RISC hardware in the business.
>
> Tried UltraSPARC IV lately? Next!

	...just another fanboy.

>
> And, it is irrelevant whether Solaris runs on SPARC/RISC or not, since 
> it also runs on Pentium4 and Opteron, which are both CISC/RISC hybrids.

	This little comment here only demonstrates that you are some sort
of demented wannabe, not even a genuine (as a matter of experience) fanboy.

	JEDIDIAH, SCSA

-- 
	The best OS in the world is ultimately useless         |||
	if it is controlled by a Tramiel, Jobs or Gates.      / | \
                                                     

On Mon, 13 Jun 2005 19:20:06 -0500, JEDIDIAH wrote:

> On 2005-06-13, lqualig@uku.co.uk <lqualig@uku.co.uk> wrote:
>>>> The same reason why "everyone" is using Microsoft Windows on their pc's?
>>
>>> No, I don't think Linux comes pre-installed on supercomputers....
>>
>>
>> That's funny - and with just the right amount of sarcasm. This would
>> make for a great Far Side or Dilbert cartoon.
>>
> 
> 	You can get prebuilt computing clusters with Linux intalled actually...

Ok, ok, but how large are those clusters? Also, the other part of my
original point still stands. No arm twisting, extorting, evil empire
keeping Linux dominant in super-computers.

Liam Slider wrote:
> On Mon, 13 Jun 2005 09:01:49 -0700, lqualig wrote:
> 
> 
>>>>The same reason why "everyone" is using Microsoft Windows on their pc's?
>>
>>>No, I don't think Linux comes pre-installed on supercomputers....
>>
>>
>>That's funny - and with just the right amount of sarcasm. This would
>>make for a great Far Side or Dilbert cartoon.
> 
> 
> I was thinking more along the lines of User Friendly myself...
> 

Frame 1: Dogbert: "I've discovered a way to make office staff more pliable."
Frame 2: Dilbert: "How?"
Frame 3: Dogbert: "I'm gonna have the ageing UNIX system ripped out..." 
[Dilbert looks interested]
Frame 4: Dogbert: "...And install Windows 2000 and Solitaire!" [Dilbert 
looks terrified]

-- 
Cheers,					http://www.dotware.co.uk
Jim			  http://www.dotware-entertainment.co.uk

I'm not scared of the man who wants ten nuclear warheads.
I'm terrified of the man who who wants just one.
  - Dr. Julia Kelly "The Peacemaker"

UNIX admin wrote:
>> No shit it will run if you port it..
> 
> 
> So if you are in agreement with me, then we have no problem.
  HPUX, AIX, et al could run games or anything else for the matter if
you port them.  Your proof is superfluous.  Market support will never
rally behind Solaris for it to happen.

This is my last post on the matter.  Don't feed the trolls.

In comp.unix.solaris Kamal R. Prasad <kamalp@acm.org> wrote:


> The best publicly available source I can think of is:-
> http://febcm.club.fr/english/chronoa13.htm

> Other than that, I happened to work at IBM India and have seen src code
> with a multi-year copyright + license to use from Bull to IBM. I can
> conlude safely that quite a bit if not all of AIX src code was
> developed by Bull and is still being maintained under contract by some
> appendage of Bull.

All you can conclude is that Bull contributed *some* to AIX code
under the agreement by which IBM provided AIX and RISC6000 tech
to Bull.

BTW, I worked for a short time at IBM's AIX Development group in
Austin (RISC6000 division HQ), working directly with the source
and builds. I can assure you, the vast majority of AIX development
has always been done there. 


>> it was quite the open secret that Bull used IBM's RISC/6000
>> workstations (rebadged) and IBM delivered a custom version
>> of AIX to Bull that had all references to "AIX" and "IBM"
>> replaced with "BullOS" and "Bull". Certainly there was a

> Probably not. I mean, they must have marketed it as AIX itself and Bull
> was just a vendor of computers, with IBM as a chip supplier.

IBM manufactured RISC6000 and sold it to Bull. This is well established.
And anyone woking at IBM should have known that IBM made their own
RISC6000 everything. I don't recall if they were actually manafactured
in Austin, but since the IBM PC company (also HQ'ed in Austin) factory
that built many of the IBM PC's was right across the
street, it was probably done in the same place.

>> great deal of collaboration on the part of Bull with IBM
>> to satisfy Bull's custom product needs, but it was IBM that
>> led the AIX development.
>>
> Im not sure I understand what customization Bull had to go through.
> They weren't into retrofitting computers to defence equipment or
> something like that. Their target customer base was the same that buys
> unix systems today.

Like any other major partner, Bull had customers with unique requirements,
and made their own variant computers/devices based on RISC6000 tech from
IBM, in addition to simply selling re-branded IBM equipment that was
completely the same in all but label. The whole arrangement was rather
unique since they were otherwise competitors. Bull was definitely way
more than just a reseller, but weren't remotely in the leading
development position to which you are ascribing them.

I'd be interested in a version of Solaris that would run on one of the
PCs I already have.

I've tried installing Solaris for Intel a few times, not much luck
though.  I also purchased a Sun Ultra-10 a few years ago.
Unfortunately, the machine was cheap but getting memory and peripherals
was expensive.

I'd really like to see Sun do a flawless delivery.  They make some
really great products, and I love their Sparc based servers.  I've used
them on several projects.  I used to have a Sun Workstation and
absolutely LOVED it.

Good luck to Sun and Scott McNealy.

Liam Slider wrote:
> On Mon, 13 Jun 2005 16:03:45 +0200, Trond Norbye wrote:
> 
> 
>>The same reason why "everyone" is using Microsoft Windows on their pc's?
> 
> 
> No, I don't think Linux comes pre-installed on supercomputers....
> 
> 
> Nor is some giant company using arm twisting tactics to ensure it's put on
> supercomputers.
> 
> 
> So you'd be wrong.
> 
Wrong!


Linux is the OS of the IBM Blue Gene supercomputers

Chris Barrera wrote:
> In comp.unix.solaris Kamal R. Prasad <kamalp@acm.org> wrote:
>
>
> > The best publicly available source I can think of is:-
> > http://febcm.club.fr/english/chronoa13.htm
>
> > Other than that, I happened to work at IBM India and have seen src code
> > with a multi-year copyright + license to use from Bull to IBM. I can
> > conlude safely that quite a bit if not all of AIX src code was
> > developed by Bull and is still being maintained under contract by some
> > appendage of Bull.
>
> All you can conclude is that Bull contributed *some* to AIX code
> under the agreement by which IBM provided AIX and RISC6000 tech
> to Bull.
>
> BTW, I worked for a short time at IBM's AIX Development group in
> Austin (RISC6000 division HQ), working directly with the source
> and builds. I can assure you, the vast majority of AIX development
> has always been done there.
>
I was referring to the origins of AIX -not subsequent changes.
Ive never been to IBM Austin, so cannot comment on geographical issues.

regards
-kamal

>
> >> it was quite the open secret that Bull used IBM's RISC/6000
> >> workstations (rebadged) and IBM delivered a custom version
> >> of AIX to Bull that had all references to "AIX" and "IBM"
> >> replaced with "BullOS" and "Bull". Certainly there was a
>
> > Probably not. I mean, they must have marketed it as AIX itself and Bull
> > was just a vendor of computers, with IBM as a chip supplier.
>
> IBM manufactured RISC6000 and sold it to Bull. This is well established.
> And anyone woking at IBM should have known that IBM made their own
> RISC6000 everything. I don't recall if they were actually manafactured
> in Austin, but since the IBM PC company (also HQ'ed in Austin) factory
> that built many of the IBM PC's was right across the
> street, it was probably done in the same place.
>
> >> great deal of collaboration on the part of Bull with IBM
> >> to satisfy Bull's custom product needs, but it was IBM that
> >> led the AIX development.
> >>
> > Im not sure I understand what customization Bull had to go through.
> > They weren't into retrofitting computers to defence equipment or
> > something like that. Their target customer base was the same that buys
> > unix systems today.
>
> Like any other major partner, Bull had customers with unique requirements,
> and made their own variant computers/devices based on RISC6000 tech from
> IBM, in addition to simply selling re-branded IBM equipment that was
> completely the same in all but label. The whole arrangement was rather
> unique since they were otherwise competitors. Bull was definitely way
> more than just a reseller, but weren't remotely in the leading
> development position to which you are ascribing them.

Solaris scaes to a lot more cpus than redhat linux. But the cpus it
scales to re sparch ones and opensolaris.org isn't going to make that
code public. It is likely code that rns on <8 cpus. IMHo linux is a
bloatware unix clone whose bloat/volatality will come in the way of its
popularity.

regards
-kamal
p.s -that is just my opinion and one ought not to take it blindly

Kamal R. Prasad wrote:
> Solaris scaes to a lot more cpus than redhat linux. But the cpus it
> scales to re sparch ones and opensolaris.org isn't going to make that
> code public. It is likely code that rns on <8 cpus. IMHo linux is a
> bloatware unix clone whose bloat/volatality will come in the way of its
> popularity.
> 
> regards
> -kamal
> p.s -that is just my opinion and one ought not to take it blindly
> 

I've had Linux running a dumb cluster (albeit a small one) with varying 
degrees of success, depending on the application overhead and the amount 
of RAM in each node. Simple number crunching such as you'd find in a 
fractal generator, is /very/ fast. Video rendering is a sight slower, 
but still plenty faster than running it on just the one box. Speed isn't 
an issue when rendering very large images, memory is, and that's where 
the cluster comes in very handy, too. The largest cluster I've built/run 
so far ran Debian 3.0r0 on 34 processors, ranging from K6II/400 up to 
PIII/1.2 through two 10MBit switches; that was a client cluster I had 
specified and built, and running through burnin tests I could compare it 
to encoding DVD-resolution video from CGI source at something like 7-9x 
realtime (or a dual 10GHz Xeon-type speed). Steenkin' fast, but even 
with that kind of processing power, very inefficient.
My current cluster employs on average 6-9 fulltime processors (depending 
on what I've got in, currently ranging from a dual PIII/766 to two 
PIIIm/1.4's and various odds and sods in between), with the resources to 
task another 11. Again, with the overheads in software, runtimes etc, 
and depending on the application, it can be very inefficient. I need a 
mainframe. :\

-- 
Cheers,					http://www.dotware.co.uk
Jim			  http://www.dotware-entertainment.co.uk

I'm not scared of the man who wants ten nuclear warheads.
I'm terrified of the man who who wants just one.
  - Dr. Julia Kelly "The Peacemaker"

Chris Barrera wrote:
> BTW, I worked for a short time at IBM's AIX Development group in
> Austin (RISC6000 division HQ), working directly with the source
> and builds. I can assure you, the vast majority of AIX development
> has always been done there. 

Since I live about a 10 minute drive from "the pink buildings" (i.e.
IBM Austin), I can confirm that lots of AIX development goes on there,
and they've been doing development on AIX there for at least 10 years.
I've known a number of people who have worked there on various
AIX-related projects.

I believe they also have done some of their Linux stuff there, although
that I'm less sure of.

   - Logan

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Sun, 12 Jun 2005 13:09:39 +0200,
 UNIX admin <tripivceta@hotmail.com> wrote:
> Dan Espen wrote:
>> Here are a number of things I might hold against Solaris:
>> 
>> 1- Recognized name brand
>> 2- officially certified
>> 3- commercial grade
>> 
>> Those are the kind of things that appeal to corporate types,
>> not people that appreciate fine operating systems.
>
> If you are implying that Linux is a fine operating system, I do emplore 
> you to delve deeper into operating systems, so that one day you may be 
> able to tell the difference. Knowledge is power, as they say.
>
> You do realize that the rest of your remarks against Solaris are not 
> based on technical merit, don't you?
>


It'd be nice if you were to come up with some technical reasons for
dissing Linux, rather than you vague innuendo and handwaving. 

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQFCrqf7d90bcYOAWPYRAjlvAKCXlkvjrpqeYOVabB9FglQ0Wr0EmACg8Ctx
hVEdYkCQmv226hPamwp98LA=
=bhlZ
-----END PGP SIGNATURE-----

-- 
Jim Richardson     http://www.eskimo.com/~warlock
Fornication, n.:
        Term used by people who don't have anybody to screw with.

Martin Etteldorf wrote:
> In comp.unix.aix UNIX admin <tripivceta@hotmail.com> wrote:
> > It turns out that not only is the pSeries platform cost prohibitive (the
> > cheapest model was close to $3000 USD)
>
> Well, I'd say it's reasonably priced. A comparably powerful x86 or
> sparc machine won't be cheaper.
>
> > but on top of that, I'd have to
> > shell out the dough for an AIX license to boot, which is also not cheap,
> > to say the least.
>
> Nonsense. Pay 35 bucks for the Media kit. Any RS/6000 comes with a
> valid AIX license.
>
hat he means is perhaps the openpower blade system running linux. I
think its called J20 and its performance sucks.

> > Meanwhile, Ultra Enterprise450, 220R, 250R, and 280Rs are selling for
> > $400-$600 USD on ebay,
>
> You're comparing a new pSeries machine with an obsolete Ultra-II
> machine? Get a life.
>
Sun beats IBM hands down in terms of price/performance. Not a chance
you can get an IBM server for the same price as an equivalent
MFLOPS/MIPS sun server.
But if you want better/faster/scalable hw -IBM provides a better deal
thanks to the technology in POWER processors.


> > and I can get Solaris just with a few clicks,
> > hassle free.
>
> I wouldn't say it's too much of a hassle calling IBM or ordering
> an AIX media kit online. Just one phonecall or just "a few clicks".
>
Dealing with IBM staff is like dealing with a govt beaurocracy.
They wouldn't want your order unless you represent a big customer.

> > There goes my idea of doing more AIX. Thanks IBM, nice job!
>
They aren't interested in dealing with low-margin home pc users. Even
if you want to setup a server at your end. you need to plae a big order
for the staff to overcome their inertia. So save your thanks for
another day.

regards
-kamal

> Go buy a 44P or alike. They are more comparable to an old U-II
> machine in both price and performance.
> But then, why am I talking to a troll anyway?
>
>
>
> 		Martin.
> -- 
> Remember to always speak softly.
> But carry a big stick!

On 14 Jun 2005 00:17:15 -0700
"Kamal R. Prasad" <kamalp@acm.org> wrote:

> I was referring to the origins of AIX -not subsequent changes.
> Ive never been to IBM Austin, so cannot comment on geographical issues.

The first version of AIX ran on IBM's PC/RT from 1986, long
before the advent of the RS/6000, and the collaboration/OEM
agreement with Bull (Bull did design a multi-processor machine,
IIRC, and if so, probably did the necessary kernel enhancements).

But the idea that Bull is the original author of AIX is incorrect.

Take care,

-- 
Stefaan
-- 
As complexity rises, precise statements lose meaning,
and meaningful statements lose precision. -- Lotfi Zadeh 

begin  In <42ad49bb$0$38039$5a6aecb4@news.aaisp.net.uk>, on 06/13/2005
   at 08:54 AM, andrew@cucumber.demon.co.uk (Andrew Gabriel) said:

>I don't know what involvement Bull had, but AIX was first developed
>by Interactive Systems Corporation under contract from IBM, as the
>reference port of SVR3.1 (IIRC) to one of IBM's systems (I guess the
>RS6000, but I don't know that for sure).

Probably the PC/RT. Note that at one time there were IX and AIX
versions for Intel and S/370 platforms, but I don't know how much of
the code was shared. IBM eventually dropped the mainframe IX and AIX
versions in favor of MVS OpenEdition[1] and VM OpenEdition[1], their
extensions to MVS and VM to support Unix applications.

[1] The names have changed several times since then.

-- 
Shmuel (Seymour J.) Metz, SysProg and JOAT  <http://patriot.net/~shmuel>

Unsolicited bulk E-mail subject to legal action.  I reserve the
right to publicly post or ridicule any abusive E-mail.  Reply to
domain Patriot dot net user shmuel+news to contact me.  Do not
reply to spamtrap@library.lspace.org

In article <20050614122632.3372eb37.tengo@deletemeecc.lu>,
	Stefaan A Eeckels <tengo@DELETEMEecc.lu> writes:
> On 14 Jun 2005 00:17:15 -0700
> "Kamal R. Prasad" <kamalp@acm.org> wrote:
> 
>> I was referring to the origins of AIX -not subsequent changes.
>> Ive never been to IBM Austin, so cannot comment on geographical issues.
> 
> The first version of AIX ran on IBM's PC/RT from 1986, long
> before the advent of the RS/6000, and the collaboration/OEM
> agreement with Bull (Bull did design a multi-processor machine,
> IIRC, and if so, probably did the necessary kernel enhancements).

Ah, that makes even more sense given the first AIX was done
by Interactive Systems Corp under contract from IBM, as ISC
were unix on x86 specialists.

> But the idea that Bull is the original author of AIX is incorrect.

Indeed.

-- 
Andrew Gabriel

On Tue, 14 Jun 2005 07:24:07 +0200, jgk wrote:

> Wrong!
> 
> 
> Linux is the OS of the IBM Blue Gene supercomputers

And IBM is forcing everyone to install Linux on all supercomputers?

In comp.unix.solaris Scott Howard <scott@hunterlink.net.au> wrote:
> In comp.unix.solaris Mike Cox <mikecoxlinux@yahoo.com> wrote:
>> But not any longer, as OpenSolaris has been confirmed to be released 
>> in a weeks time from June 8th!  OpenSolaris, SchilliX, and other 
> 
> Confirmed? By who? The lady who works at the corner shop down the road?
> I'll put _serious_ money on that date being wrong.

And as predicted, it was wrong...

6 days later, and OpenSolaris is now available from http://opensolaris.org/ !!

(OK, so it was almost a week, but not quite :)

  Scott

On Tue, 14 Jun 2005, Kamal R. Prasad wrote:

> Solaris scaes to a lot more cpus than redhat linux. But the cpus it
> scales to re sparch ones and opensolaris.org isn't going to make that
> code public. It is likely code that rns on <8 cpus. IMHo linux is a

You couldn't be more wrong, as a look at www.opensolaris.org will
confirm.

-- 
Rich Teer, SCNA, SCSA, OpenSolaris CAB member

President,
Rite Online Inc.

Voice: +1 (250) 979-1638
URL: http://www.rite-group.com/rich

aixdude@yahoo.com wrote:

> I suggest moving the thread to an advocacy group, where it belongs.
> 

This is an advocacy group.

So get your thread the hell out of here.

Mike Cox wrote:
> Is Joerg Schilling the next Linus Torvalds?  Joerg has advantages that
> Linus never had.  25 years of Unix engineering expeince.  And a
> codebase battle hardened by professional Ivy league engineers.  Unlike
> Linux, SchilliX, Joerg's OpenSolaris distro will support the proper
> SCSI commands.  It won't be just a kernel.
>
> AIX is long in the tooth compared to Solaris 10.   AIX lacks DTrace,
> Zones and other goodies.  Linux doesn't have the stability or the
> features of Solaris either, but it was open source.
>
> But not any longer, as OpenSolaris has been confirmed to be released
> in a weeks time from June 8th!  OpenSolaris, SchilliX, and other
> OpenSolaris distros go head to head with Linux.  I'll bet the
> superiour offering will kick some serious linux (and AIX) butt.
>
> http://www.osdir.com/Article5883.phtml

OK, I am now sending the 157th response to this post. I don't know
about the rest of you, but I've had enough of the pissing contest over
operating systems.

I have noticed a lot of new names posting to the group. It figures,
long on opinion, short on answers to real every day problems admins
face.

I suggest moving the thread to an advocacy group, where it belongs.

On 14 Jun 2005 11:27:28 GMT
andrew@a17 (Andrew Gabriel) wrote:

> In article <20050614122632.3372eb37.tengo@deletemeecc.lu>,
> 	Stefaan A Eeckels <tengo@DELETEMEecc.lu> writes:
> > On 14 Jun 2005 00:17:15 -0700
> > "Kamal R. Prasad" <kamalp@acm.org> wrote:
> > 
> >> I was referring to the origins of AIX -not subsequent changes.
> >> Ive never been to IBM Austin, so cannot comment on geographical issues.
> > 
> > The first version of AIX ran on IBM's PC/RT from 1986, long
> > before the advent of the RS/6000, and the collaboration/OEM
> > agreement with Bull (Bull did design a multi-processor machine,
> > IIRC, and if so, probably did the necessary kernel enhancements).
> 
> Ah, that makes even more sense given the first AIX was done
> by Interactive Systems Corp under contract from IBM, as ISC
> were unix on x86 specialists.

Actually, the PC/RT (or RT PC, it stood for RISC Technology
Personal Computer) wasn't based on the x86 but on a
precursor of the POWER microprocessor. I don't think
ISC did the original implementation, but they were involved
in porting some of the Berkeley extensions to AIX 1 and 2. 

I remember distinctly how the IBM sales critter sang the
praises of AIX, especially in the user interaction department,
where he said that IBM had taken special care to provide
elaborate, IBM standard type error messages instead of the
usual "cryptic and unprofessional Unix messages" (I guess he
meant "lp on fire").

Somehow I got the impression that a lot of AIX was if not
rewritten by IBM, at least heavily customised. 

Take care,

-- 
Stefaan
-- 
As complexity rises, precise statements lose meaning,
and meaningful statements lose precision. -- Lotfi Zadeh 

In comp.os.linux.advocacy, jabailo@texeme.com
<jabailo@texeme.com>
 wrote
on Tue, 14 Jun 2005 09:14:28 -0700
<42AF0264.9030400@texeme.com>:
> aixdude@yahoo.com wrote:
>
>> I suggest moving the thread to an advocacy group, where it belongs.
>> 
>
> This is an advocacy group.
>
> So get your thread the hell out of here.

Ah..hem.

  ** Newsgroups: comp.unix.solaris,comp.os.linux.advocacy,comp.unix.aix **

OK, so you're right for 1 out of the three, but the other two
don't look like advocacy groups to me. :-)

Followups reset.

-- 
#191, ewill3@earthlink.net
It's still legal to go .sigless.

Martin Etteldorf <etteldor@mail.ru> wrote:

> >>False. It runs on PPC and POWER only, but not limited to IBM.
> > 
> > Show me AIX running on a PowerMAC,
> 
> Apple Network Server 500/700

not a powermac - it doesn't run any kind of Mac OS.

-- 
Morten Reippuert Knudsen :-)                <http://blog.reippuert.dk>

PowerMac G5: 1.6GHz, 1.25GB RAM, 300+300GB SATA, 8xDVD+/-RW, Bluetooth
mus+tastatur, R9600Pro, iSight, eyeTV200 & LaCie Photon18Vision TFT

begin  In <1118765792.087946.78990@g44g2000cwa.googlegroups.com>, on
06/14/2005
   at 09:16 AM, aixdude@yahoo.com said:

>Newsgroups: comp.unix.solaris,comp.os.linux.advocacy,comp.unix.aix

>I suggest moving the thread to an advocacy group, where it belongs.

See above ;-)

The OP was clearly a troll trying to trigger a flame war with his
inappropriate crossposting.

-- 
Shmuel (Seymour J.) Metz, SysProg and JOAT  <http://patriot.net/~shmuel>

Unsolicited bulk E-mail subject to legal action.  I reserve the
right to publicly post or ridicule any abusive E-mail.  Reply to
domain Patriot dot net user shmuel+news to contact me.  Do not
reply to spamtrap@library.lspace.org

Stefaan A Eeckels wrote:
> I remember distinctly how the IBM sales critter sang the
> praises of AIX, especially in the user interaction department,
> where he said that IBM had taken special care to provide
> elaborate, IBM standard type error messages instead of the
> usual "cryptic and unprofessional Unix messages" (I guess he
> meant "lp on fire").

Oh, you mean like the ^%$#-ing "add a filesystem" menu option
in SMIT that contributed to my wiping a filesystem with
important data on it?  It was back when I was just starting
to admin AIX systems, and an RS/6000 had died.  The user was
throwing a fit[1] about how they needed access to their data NOW,
so I grabbed the external disk and plugged it in to another
RS/6000 nearby.

I wasn't sure how to import data that had been on another
AIX system, so I looked through the menus to find something
appropriate.  Eventually, I found "add a filesystem", and it
seemed perfect:  I thought, hey, I *have* this filesystem
right here on this disk, and what I want to do is *add* it to
the list of filesystems that the AIX system knows!  In other
words, I have this filesystem, and I want to add it to
/etc/filesystems.  Perfect.

So, I selected that, and told it to go ahead, and it seemed
to be taking a while.  After I tried it two or three times,
I felt something was wrong.  And something was.  Because in
IBM-speak, "add" means "create" (as in mkfs), and you're
supposed to just know that it doesn't mean any of the other
things you'd find under "add" in the dictionary.  (If you
work with smit long enough, you'll eventually infer that
"add" means "create" specifically.  Hopefully without
mkfs-ing over the source code someone just wrote.)

The point is that I *know* this must've happened because
someone at IBM had a set of standards for how AIX should
be, and one of those standards was that the menu items in
smit should be nice and uniform, regardless of whether that
was actually clear or was misleading in some cases.

   - Logan

[1]  We were under a very tight schedule, so that was understandable.
      Unfortunately, this was before I learned a critical skill
      of being a sysadmin:  let the users vociferously demand
      whatever they want, extract the facts and take them as
      input, ignore the other 99%, and then do what *you* know to
      be the reasonable thing to do.

Jim Richardson wrote:
> It'd be nice if you were to come up with some technical reasons for
> dissing Linux, rather than you vague innuendo and handwaving. 

* You can't just build a universal kernel that contains modular
   support for everything in the source code, which is demand
   loaded as the hardware is found.  Instead, you have to customize
   the kernel for every machine or live with bloat.  This is not
   so wonderful if you have hundreds of machines, some of which
   are the same as each other and some of which are not.

* Users have to keep up to date on which release kernels are
   too buggy to run without crashing the machine.  For example,
   2.6.7 was OK, but one of the ones following that (2.6.10, I
   think) was horribly unstable on some pretty mainstream
   hardware and not to be trusted.

* On all the distributions I've ever seen, there is apparently
   no concept of a stable release level at which the code is
   branched so that you can stay at this level and apply only
   bug fixes.  Instead, the solution to bugs is to upgrade to
   the latest release of whatever subsystem has the bug.  This
   is not good at for system stability because it means you
   have two choices:  install new bugs along with your bugfixes,
   or install nothing.  Meanwhile, some other operating systems
   have patches available that do nothing but fix bugs and don't
   try to introduce any new features or other updates.

Those are the top 3 most annoying ones for me, personally.

   - Logan

In comp.os.linux.advocacy Logan Shaw <lshaw-usenet@austin.rr.com> wrote:
> * You can't just build a universal kernel that contains modular
>   support for everything in the source code, which is demand
>   loaded as the hardware is found. 

Knoppix is doing a very good job at this.

> * On all the distributions I've ever seen, there is apparently
>   no concept of a stable release level at which the code is
>   branched so that you can stay at this level and apply only
>   bug fixes.  Instead, the solution to bugs is to upgrade to
>   the latest release of whatever subsystem has the bug.

So you have never seen Debian/stable.

Daniel Tryba wrote:
> In comp.os.linux.advocacy Logan Shaw <lshaw-usenet@austin.rr.com> wrote:
> 
>>* You can't just build a universal kernel that contains modular
>>  support for everything in the source code, which is demand
>>  loaded as the hardware is found. 

> Knoppix is doing a very good job at this.

That's good, but isn't Knoppix only a Live CD distribution?  That
means it isn't really useful for regular systems on which you want
to have an operating system that boots off the hard disk and
can be customized, etc., right?

>>* On all the distributions I've ever seen, there is apparently
>>  no concept of a stable release level at which the code is
>>  branched so that you can stay at this level and apply only
>>  bug fixes.  Instead, the solution to bugs is to upgrade to
>>  the latest release of whatever subsystem has the bug.

> So you have never seen Debian/stable.

When I checked out Debian and installed it as a test, it seemed
questionable whether it was a viable distribution anymore.
Maybe with the 3.0 and 3.1 updates they've finally released,
that has changed.  I also found dselect to be painful and
difficult to use (keep in mind I prefer the command line and
use "vi" as my text editor, so it's not as if I can't cope with
curses apps), but that's just personal preference maybe.

   - Logan

In comp.os.linux.advocacy Logan Shaw <lshaw-usenet@austin.rr.com> wrote:
>> Knoppix is doing a very good job at this.
> 
> That's good, but isn't Knoppix only a Live CD distribution?

No.

>> So you have never seen Debian/stable.
> 
> When I checked out Debian and installed it as a test, it seemed
> questionable whether it was a viable distribution anymore.

Now what do you want? Stable is extremly stable, AFAIK the only upgraded
package in 3.0 was ssh (updated after some panic attack from Theo (which
proved unnecessary since that version wasn't vulnerable IIRC)) during
it's 3 year lifecycle

> Maybe with the 3.0 and 3.1 updates they've finally released,
> that has changed.  I also found dselect to be painful and
> difficult to use

dselect could be avoided in 3.0 stable if you wanted (and should be
since dselect is a really scary prog).

> (keep in mind I prefer the command line and use "vi" as my text
> editor, so it's not as if I can't cope with curses apps), but that's
> just personal preference maybe.

So what kept you from using it? vi is installed by default, and Debian
doesn't force you to use frontends (there is a readline alternative :).

Daniel Tryba wrote:
> In comp.os.linux.advocacy Logan Shaw <lshaw-usenet@austin.rr.com> wrote:
> 
>>>Knoppix is doing a very good job at this.
>>
>>That's good, but isn't Knoppix only a Live CD distribution?

> No.

Then why does www.knoppix.org, the first hit when I type "knoppix"
in on google.com, say that "KNOPPIX is a bootable CD" on a web
page whose title is "KNOPPIX Linux Live CD"?  If it's more than a
Live CD distribution, they are doing a good job hiding it.  Nothing
on that page indicates anything otherwise, and it all strongly
implies CD-only.  That would tend to make me believe one of two
things:

1)  Knoppix is only a Live CD, or
2)  Knoppix can be installed on the hard disk, but the web
     page is badly written and highly misleading.

I don't really like either one...

>>>So you have never seen Debian/stable.
>>
>>When I checked out Debian and installed it as a test, it seemed
>>questionable whether it was a viable distribution anymore.
> 
> 
> Now what do you want? Stable is extremly stable, AFAIK the only upgraded
> package in 3.0 was ssh (updated after some panic attack from Theo (which
> proved unnecessary since that version wasn't vulnerable IIRC)) during
> it's 3 year lifecycle

I last checked out Debian before 3.0 was released.  Things may be
different now that 3.0 and 3.1 are out there, but at the time I
checked it out, when I looked at the Debian stuff, I felt like
tumbleweeds were going to start rolling across my field of vision
at any moment.

Looking at the news section on debian.org, it would seem that
that must've been about 3 or 3.5 years ago, and Debian isn't dead
yet, so maybe it has been revitalized, or maybe the rumors of its
death were greatly exaggerated from the very beginning.

At any rate, to me a kernel that doesn't have to be manually
configured and rebuilt (unless you are doing something severely
out of the ordinary and experimental) and a stable branch with
only bugfixes are BOTH important features for an operating system.
With Linux, it seems like I can have one or the other but I can't
have both.  One reason I like Solaris is that it already has both
of those things worked out.

   - Logan

Logan Shaw wrote:

> Daniel Tryba wrote:
>> Logan Shaw <lshaw-usenet@austin.rr.com>

>>>That's good, but isn't Knoppix only a Live CD distribution?
> 
>> No.
> 
> Then why does www.knoppix.org, the first hit when I type "knoppix"
> in on google.com, say that "KNOPPIX is a bootable CD" on a web
> page whose title is "KNOPPIX Linux Live CD"? 

<URL:http://www.knoppix.net/wiki/Hd_Install_HowTo>
'You CAN install newer versions of Knoppix on harddisk, if you wish,
though installation is not required for productive use. Knoppix is
designed to be a GNU/Linux LiveCD and not a HD installed Linux distro.'


-- 
Timo Pirinen
pirisisi@dlc.fi

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Wed, 15 Jun 2005 06:21:01 GMT,
 Logan Shaw <lshaw-usenet@austin.rr.com> wrote:
> Jim Richardson wrote:
>> It'd be nice if you were to come up with some technical reasons for
>> dissing Linux, rather than you vague innuendo and handwaving. 
>
> * You can't just build a universal kernel that contains modular
>    support for everything in the source code, which is demand
>    loaded as the hardware is found.  Instead, you have to customize
>    the kernel for every machine or live with bloat.  This is not
>    so wonderful if you have hundreds of machines, some of which
>    are the same as each other and some of which are not.
>

Actually, yes you can. In fact, this is what most of the distro's
install image does. Using a combination of a well thought out initrd
and a modular kernel. For example, I am using the same kernel/install
image, on three different machines, with different chipsets, video, and
ancillary componants. (Ubuntu Hoary in this case) 

> * Users have to keep up to date on which release kernels are
>    too buggy to run without crashing the machine.  For example,
>    2.6.7 was OK, but one of the ones following that (2.6.10, I
>    think) was horribly unstable on some pretty mainstream
>    hardware and not to be trusted.


That's what the distros do. While there are certainly mistakes made, the
distros vet the kernel (and other software) for stability, and security.
If you want to roll your own, then you are on your own, which is how it
should be. 


>
> * On all the distributions I've ever seen, there is apparently
>    no concept of a stable release level at which the code is
>    branched so that you can stay at this level and apply only
>    bug fixes.  Instead, the solution to bugs is to upgrade to
>    the latest release of whatever subsystem has the bug.  This
>    is not good at for system stability because it means you
>    have two choices:  install new bugs along with your bugfixes,
>    or install nothing.  Meanwhile, some other operating systems
>    have patches available that do nothing but fix bugs and don't
>    try to introduce any new features or other updates.

You have either not looked at  many distros, or are being flippant. 


 Many distros do exactly this. Of the ones I am familiar with, Debian,
Ubuntu, Redhat, Novel/Suse all maintain a stable distro, for a given
period of time (or until the next one is done, like Debian, which may
take 3 years...)

Ubuntu, after they release a version, maintain updates (not upgrades,
updates) to the previous version for 18months. So you have 24months of
update support, including security fixes, and major bugfixes, but not,
unless absolutely neccessary, new versions. Of course, if you want, you
can upgrade to newer versions yourself, or step onto the testing distro,
but that is your choice.  



>
> Those are the top 3 most annoying ones for me, personally.

:) Three down....

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQFCsAQGd90bcYOAWPYRAkFtAJ9FQ7Vl76isiU0ec02fWy9dO9TcNQCgoMwA
xMPzVw1Z/jH2n3SICG8zW1s=
=qL0Q
-----END PGP SIGNATURE-----

-- 
Jim Richardson     http://www.eskimo.com/~warlock
I have plenty of talent and vision. I just don't give a
damn.

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Wed, 15 Jun 2005 08:24:20 GMT,
 Logan Shaw <lshaw-usenet@austin.rr.com> wrote:
> Daniel Tryba wrote:
>> In comp.os.linux.advocacy Logan Shaw <lshaw-usenet@austin.rr.com> wrote:
>> 
>>>* You can't just build a universal kernel that contains modular
>>>  support for everything in the source code, which is demand
>>>  loaded as the hardware is found. 
>
>> Knoppix is doing a very good job at this.
>
> That's good, but isn't Knoppix only a Live CD distribution?  That
> means it isn't really useful for regular systems on which you want
> to have an operating system that boots off the hard disk and
> can be customized, etc., right?
>

No, you can install Knoppix, you can also use the CD image (either from
the CD, or an iso file on the HD) with a partition (or USB thumbdrive,
etc) for config changes, updates to software, and personal files. 

Take the CD and the thumbdrive to another machine, and your "system"
comes with you. I have been doing this for several years now. Works
great. Go to a client's office, pop the CD and the thumbdrive into an
unused machine, boot up, and have the tools I am used to working with,
right there. No need to scour the net for them, or ask "can I install
appfoo on this machine?" 

As a side benefit, I have found that Knoppix, when used on a machine
with sufficient RAM (about 512 is good) it is generally faster than the
installed version of MS-Windows. Since Linux has good caching
algorithms. Mind you, Knoppix will run on machines with far less RAM,
but definately benefits from more RAM.

You can do much the same thing with Kanotix, Mepis, and PCLinuxOS to
name but a few. For more info on livecds, of many flavours and uses,
check out <http://www.frozentech.com/content/livecd.php>. There are
distros for computer forensics, multimedia, rescue, small footprint
systems, security, games, scientific analysis. &tc. 

>>>* On all the distributions I've ever seen, there is apparently
>>>  no concept of a stable release level at which the code is
>>>  branched so that you can stay at this level and apply only
>>>  bug fixes.  Instead, the solution to bugs is to upgrade to
>>>  the latest release of whatever subsystem has the bug.
>
>> So you have never seen Debian/stable.
>
> When I checked out Debian and installed it as a test, it seemed
> questionable whether it was a viable distribution anymore.
> Maybe with the 3.0 and 3.1 updates they've finally released,
> that has changed.  I also found dselect to be painful and
> difficult to use (keep in mind I prefer the command line and
> use "vi" as my text editor, so it's not as if I can't cope with
> curses apps), but that's just personal preference maybe.

Don't use dselect then, use apt. I have never used dselect beyond once
or twice, I agree, too painful!

However, Debian is exactly what you were asking for. The stable branch
is exactly that. Stable. The only reason for new versions of stuff, is
major bugfixes that can't be backported to the provided version. It took
3 years for Debian testing to become Debian Stable, and in that time,
the old Stable (Woody) saw no major revision changes of any apps I am
aware of. There are almost certainly some, since there were some 16,000
packages available in Woody, but the vast majority were stable in the
sense of not changing, as well as not crashing. 


For something a little more "predictable" in the release department,
Ubuntu does 6month releases, supporting each release for 24 months.
Redhat also release a version, (RHEL 3 for example) and make only
security and bugfixes to it. Not version changes, They aren't quite as
strict as Debian I think, but they certainly fit what you seem to be
seeking. 


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQFCsAbcd90bcYOAWPYRAu1sAJ95GQEwTRUgyfJviafST/YLI0ujkACgpLSw
lJBn/EFIDXFJFuKsGSi7LUo=
=semT
-----END PGP SIGNATURE-----

-- 
Jim Richardson     http://www.eskimo.com/~warlock
"I think quotes are very dangerous things."
	-- Kate Bush

Logan Shaw wrote:

> * You can't just build a universal kernel that contains modular
>    support for everything in the source code, which is demand
>    loaded as the hardware is found.  Instead, you have to customize
>    the kernel for every machine or live with bloat.  This is not
>    so wonderful if you have hundreds of machines, some of which
>    are the same as each other and some of which are not.

This is not true. I will speak only about the debian kernel because it is
the one I know best, but it is more general. You can chose for most
functionalities whether you want them in the kernel or as modules, and
distributions usually put plenty of things in modules, and then have
programs to load the apropriate modules for your machine. There are still
some cases where your hardware needs a recompile of the kernel, but then
there are cases where solaris_x86 won't work, you have to check the HCL
for your distro first.

PS: I am not trying to say linux is better, I prefer solaris, I just want
people to chose it for right reasons.

On 2005-06-15, Logan Shaw <lshaw-usenet@austin.rr.com> wrote:
> Daniel Tryba wrote:
>> In comp.os.linux.advocacy Logan Shaw <lshaw-usenet@austin.rr.com> wrote:
[deletia]
>>>When I checked out Debian and installed it as a test, it seemed
>>>questionable whether it was a viable distribution anymore.
>> 
>> 
>> Now what do you want? Stable is extremly stable, AFAIK the only upgraded
>> package in 3.0 was ssh (updated after some panic attack from Theo (which
>> proved unnecessary since that version wasn't vulnerable IIRC)) during
>> it's 3 year lifecycle

	Make up your damn mind. Either you want it to be a stable
configuration or mangled mess living on the bleeding edge.

>
> I last checked out Debian before 3.0 was released.  Things may be
> different now that 3.0 and 3.1 are out there, but at the time I
> checked it out, when I looked at the Debian stuff, I felt like
> tumbleweeds were going to start rolling across my field of vision
> at any moment.

[deletia]

-- 
	The best OS in the world is ultimately useless         |||
	if it is controlled by a Tramiel, Jobs or Gates.      / | \
                                                     

Marc <marc.glisse@gmail.com> writes:

> Logan Shaw wrote:
>
>> * You can't just build a universal kernel that contains modular
>>    support for everything in the source code, which is demand
>>    loaded as the hardware is found.  Instead, you have to customize
>>    the kernel for every machine or live with bloat.  This is not
>>    so wonderful if you have hundreds of machines, some of which
>>    are the same as each other and some of which are not.
>
> This is not true. I will speak only about the debian kernel because it is
> the one I know best, but it is more general. You can chose for most
> functionalities whether you want them in the kernel or as modules, and
> distributions usually put plenty of things in modules, and then have
> programs to load the apropriate modules for your machine. There are still
> some cases where your hardware needs a recompile of the kernel, but then
> there are cases where solaris_x86 won't work, you have to check the HCL
> for your distro first.

But e.g. on Linux you have to (AFAIK) compile in module for your root FS
(be it ext2, ext3 or whatever), on Solaris you don't.

E.g., on Solaris 10 GA, here is the list of first 10 modules loaded:

% uname -a
SunOS dc933 5.10 Generic_118844-01 i86pc i386 i86pc
% modinfo | head -10
 Id Loadaddr   Size Info Rev Module Name
  0 fe800000  94d78   -   0  unix ()
  1 fe8643f0   eec6   -   0  krtld ()
  2 fe86f2e0 136068   -   0  genunix ()
  4 fe96f000   3b38   1   1  specfs (filesystem for specfs)
  6 fe9729a8   34a0  16   1  devfs (devices filesystem 1.13)
  8 fe976d14   2aa4   1   1  TS (time sharing sched class)
  9 fe979318    87c   -   1  TS_DPTBL (Time sharing dispatch table)
 11 fe979374    ebc   -   1  pci_autoconfig (PCI BIOS interface 1.41)
 12 fe97a198  2d3f0   2   1  ufs (filesystem for ufs)

Here is the similar list on a latest Solaris Express:

% uname -a
SunOS lokrum 5.11 snv_16 i86pc i386 i86pc
% modinfo | head -22
 Id Loadaddr   Size Info Rev Module Name
  0 fe800000  adfe0   -   0  unix ()
  1 fe86e1f0  128c6   -   0  krtld ()
  2 fe87c118 13cde0   -   0  genunix ()
  4 fe983000   3b08   1   1  specfs (filesystem for specfs)
  5 f9e06000  12510 155   1  dtrace (Dynamic Tracing)
  6 fe986978   34a8  16   1  devfs (devices filesystem 1.13)
  7 fe989b80   2aa4   1   1  TS (time sharing sched class)
  8 fe98c184    87c   -   1  TS_DPTBL (Time sharing dispatch table)
 10 fe98c1e0   51c0   -   1  pci_autoconfig (PCI BIOS interface 1.43)
 11 fe9904a8   21bc   1   1  uppc (UniProcessor PC)
 12 fe9924b4  228d8   -   1  acpica (ACPI interpreter 1.1)
 14 fe9baae0   3354   1   1  rootnex (i86pc root nexus 1.131)
 15 fe9bdc5c    124  57   1  options (options driver)
 16 fe9bdd10    7bc   2   1  pseudo (nexus driver for 'pseudo' 1.27)
 17 fe9be334    464  11   1  clone (Clone Pseudodriver 'clone')
 18 fe9be648   abec 174   1  scsi_vhci (SCSI VHCI Driver 1.37)
 19 fe9c8554   bfb8   -   1  scsi (SCSI Bus Utility Routines)
 20 fe9cd72c   17c8 120   1  isa (isa nexus driver for 'ISA' 1.36)
 21 fe9ceb4c   1a54   -   1  busra (Bus Resource Allocator (BUSRA) )
 22 fe9b398c   1054  12   1  sad (STREAMS Administrative Driver ')
 23 fe9d0330  2d710   2   1  ufs (filesystem for ufs)

Bye, Dragan

-- 
Dragan Cvetkovic, 

To be or not to be is true. G. Boole      No it isn't.  L. E. J. Brouwer

!!! Sender/From address is bogus. Use reply-to one !!!

Dragan Cvetkovic  wrote:

> But e.g. on Linux you have to (AFAIK) compile in module for your root FS
> (be it ext2, ext3 or whatever), on Solaris you don't.

No, you create an initrd. There is one that comes with the kernel in most
distributions. It includes drivers for several FS, and is used only long
enough to load the kernel and the apropriate modules. Here lsmod shows
that I have ext3 as a module, whereas it is my root FS. (Modules for all
fs not being loaded is actually even a problem when using usb keys on
linux, because the automatic detection of the fs on the key only tries fs
for which the module is already loaded, which is rarely the case for
vfat, so you have to say it explicitly (but I expect they will correct
that soon))

And I believe the situation is similar for solaris, from what I remember
of a discussion on making zfs bootable.

Disclaimer: I am no expert, all I say may be complete nonsense.

On Wed, 15 Jun 2005, JEDIDIAH wrote:

> 	Make up your damn mind. Either you want it to be a stable
> configuration or mangled mess living on the bleeding edge.

It's perfectly reasonable to want both: stable for production
servers, and bleeding edge for one's own desktop.

-- 
Rich Teer, SCNA, SCSA, OpenSolaris CAB member

President,
Rite Online Inc.

Voice: +1 (250) 979-1638
URL: http://www.rite-group.com/rich

On Wednesday 15 June 2005 00:21, Logan Shaw <lshaw-usenet@austin.rr.com>
(<hNPre.55866$6g3.4911@tornado.texas.rr.com>) wrote:

> * You can't just build a universal kernel that contains modular
>    support for everything in the source code, which is demand
>    loaded as the hardware is found.

You can just build a universal kernel, but you don't have to.

>    Instead, you have to customize 
>    the kernel for every machine or live with bloat.

You don't have to build a customized kernel, but you can.

> * Users have to keep up to date on which release kernels are
>    too buggy to run without crashing the machine.

With the exception of 2.6.8 and its broken NFS support, you're wrong again. 
(The bug was fixed in about twenty-four hours.)

>    For example, 
>    2.6.7 was OK, but one of the ones following that (2.6.10, I
>    think) was horribly unstable on some pretty mainstream 
>    hardware and not to be trusted.

You're wrong again, unless you mean some unknown, unnamed kernel versions
which can be exceptionally unstable.  I don't trust them either.

> * On all the distributions I've ever seen, there is apparently
>    no concept of a stable release level at which the code is

You're wrong again, unless you mean some unknown, unnamed distributions. 
All the unknown, unnamed distributions I've seen won't even boot.

>    Meanwhile, some other operating systems
>    have patches available that do nothing but fix bugs and don't
>    try to introduce any new features or other updates.

These unknown, unnamed operating systems have perfect patches that fix
nonexistent bugs in their perfect operating systems.

> Those are the top 3 most annoying ones for me, personally.

They're imaginary annoyances for you, personally.

Arkady Duntov <arkady-duntov@brotherhood.ua> writes:

> With the exception of 2.6.8 and its broken NFS support, you're wrong again. 
> (The bug was fixed in about twenty-four hours.)

We still have problems using Linux (running Fedora Core r3,  2.6.11 kernel) NFS
clients in NFSv4 environment. They would panic once in a while. Very bad
if that machine is your company's email server. We had to disable NFS4
completely, which is a bloody shame concerning its advantages :-(

Bye, Dragan

-- 
Dragan Cvetkovic, 

To be or not to be is true. G. Boole      No it isn't.  L. E. J. Brouwer

!!! Sender/From address is bogus. Use reply-to one !!!

On Wed, 15 Jun 2005, Dragan Cvetkovic wrote:

> Arkady Duntov <arkady-duntov@brotherhood.ua> writes:
>
> > With the exception of 2.6.8 and its broken NFS support, you're wrong again.
> > (The bug was fixed in about twenty-four hours.)
>
> We still have problems using Linux (running Fedora Core r3,  2.6.11 kernel) NFS
> clients in NFSv4 environment. They would panic once in a while. Very bad
> if that machine is your company's email server. We had to disable NFS4

Linux has broken NFS?  What else is new?!  ;-)

-- 
Rich Teer, SCNA, SCSA, OpenSolaris CAB member

President,
Rite Online Inc.

Voice: +1 (250) 979-1638
URL: http://www.rite-group.com/rich

Martin Etteldorf wrote:
> Maybe you should stay with your 32 bit machine instead if it is
> sufficient for your needs.

Hello! It's not about what I *need* (I don't *need* AIX, I have 
IRIX/IRIX64, HP-UX and Solaris for that). It's about what I *want*, and 
I *want* to learn AIX much deeper, but I'm not prepared to pay with my 
liver and kidneys just to be able to get an aging 32-bit pSeries rack 
mountable server.

> A B50 for $3000? You're joking. They're selling for less than
> 500$ (or 400 Euros, for that matter) these days on ebay.

Did you happen to look on ebay for a B50? Interestingly enough, good 
luck finding one.

Kevin Bowling wrote:

> You are comparing apples to oranges in Price/Performance ratios here.
> In addition, lack of newer (and older, even MCA) RS/6000s on eBay is
> beacuse they are still doing their JOB, not needing to be replaced every
> couple years like other garbage.

I don't think so! IBMs may be fast, but they're nothing so special when 
compared to their other UNIX counterparts. You can't be telling me that 
a pSeries is faster (or slower, for that matter) from a SunFire V210 or 
a V20z, or a 280R, or an SGI Origin 300, and so on. That's not realistic.

Rich Teer wrote:

> FWIW, I'm typing this on a Acer Ferrari 3400 x64 laptop.  Running
> Solaris 10, of course.

Yeah yeah! Lucky you with Casper's frkit. You had the good fortune to 
meet up with him in person, the rest of us aren't so lucky.

Martin Etteldorf wrote:

> False. It runs on PPC and POWER only, but not limited to IBM.

What are you telling me here? Are you implying that I can run AIX on an 
iBook, or a G5?

Arkady Duntov wrote:

> Show me OpenSolaris with Unix 2003 certification.

What we really should show you, is the finger.
Not that you're right, your scribble here is completely without any 
technical merit whatsoever.

Ku Karlovsky wrote:

> If FUD is disallowed, you'll have to stop posting altogether.
> That would be a good thing.

Why don't you show me how that's done? You know, lead by example?

UNIX admin <tripivceta@hotmail.com> writes:

>Rich Teer wrote:

>> FWIW, I'm typing this on a Acer Ferrari 3400 x64 laptop.  Running
>> Solaris 10, of course.

>Yeah yeah! Lucky you with Casper's frkit. You had the good fortune to 
>meet up with him in person, the rest of us aren't so lucky.


We're planning to release most of it as part of OpenSOlaris; perhaps
I can check today whether I can release the stuff I personnally wrote
under the CDDL rightaway.  (It would be prototype stuff, but it
is kinda useful to have power management and battery monitoring on
your Ferrari)

Casper

Ku Karlovsky wrote:

> No, that would be Sun's OpenSolaris and their "we're the openest of
> open source and we have a Community Advisory Board who tells you so."

Just like Linux kernel developers decide for the rest of you sheep what 
you'll get to run. But since it's Linux, that makes that OK.

> OpenSolaris IS NOT UNIX BUT YOU DON'T UNDERSTAND.  The OpenGroup
> doesn't list Solaris 10 or any of its marketing derivates as
> qualified.

Looky here, OpenSolaris is a 1:1 exact copy of Solaris, which is UNIX 
certified. Solaris10 may not be explicitly certified, but since, unlike 
crappy Linux, Sun makes sure latest Solaris is backward compatible, we 
do in fact run certified UNIX.

It's just that you don't have a clue, that's the real culprit here. Now, 
troll, hop happily along... you Linux box is waiting for you to lick it 
from front to end. Don't forget to suck all the dustbunnies out in the 
process, it'll help overclocking the CPU by making it cooler. Perhaps 
you'll get to see one more sitting penguin from the "overclocking 
boost", oh joy!

> A "UNIX admin" does not an engineer make.

Hehe, troll... completely clueless, thinking he knows exactly what's 
going on. You must be a Linux user.

UNIX admin <tripivceta@hotmail.com> writes:

> Looky here, OpenSolaris is a 1:1 exact copy of Solaris, which is UNIX
> certified. Solaris10 may not be explicitly certified, but since, unlike
> crappy Linux, Sun makes sure latest Solaris is backward compatible, we do
> in fact run certified UNIX.

Actually, Solaris 10 has been certified, a few days ago this email landed
in my mailbox:

[snip]


***************************************************************************
* UNIX 03 CERTIFICATION ON 3 SIMULTANEOUS ARCHITECTURES
****************************************************************************

We are delighted to announce that Sun Microsystems has registered
Solaris 10 on their SPARC and X86 platforms as conforming to the
UNIX 03 Product Standard.

-- Solaris 10 Operating System - on 32-bit & 64-bit SPARC based systems
-- Solaris 10 Operating System - on 32-bit X86 based system
-- Solaris 10 Operating System - on 64-bit X86 based system


The Conformance Statements will be available in the next few days at
http://www.opengroup.org/csq/

For more information on UNIX 03, please refer to
http://www.unix.org/unix03.html

For the latest official list of UNIX registered products, please refer to
http://www.opengroup.org/openbrand/register/


-- 
Dragan Cvetkovic, 

To be or not to be is true. G. Boole      No it isn't.  L. E. J. Brouwer

!!! Sender/From address is bogus. Use reply-to one !!!

Liam Slider wrote:

> Ok, ok, but how large are those clusters? Also, the other part of my
> original point still stands. No arm twisting, extorting, evil empire
> keeping Linux dominant in super-computers.

Obviously you've never attended boardroom meetings where these types of 
deals are made. But I have, and let me tell you, there's "a little more" 
than just armtwisting tactics going on.

I suggest real distro like a Slackware or Debian. Both rock stable.
I can see from post above that people who complains about linux
don't know realy this system. 


-- 
regards,
Corvin

JEDIDIAH wrote:
  	
> 	...been "officially supported" in the past. Many of us still
> remember those days.

Nooo, Solaris10 on x86 is officially supported by Oracle for Oracle10g.

Hello. Check it.

> 	Larry doesn't merely "certify" Linux. It is his flagship high 
> performance platform.

For right now. But we will see where Larry ends up.

> 	What you said above doesn't have any relevance.
> 
> 	Linux is Larry's favorite son now. No amount of spin control will
> change that.

NOW. But Larry won't guarantee you it stays that way. If I know Larry, 
and I know him, he'll ditch Linux as soon as he smells blood.

> 	This is a bald faced lie on both counts?
> 
> 	Have you ever even touched a fireplane system? I doubt it.

Do tell? Perhaps I'll learn something new about supercomputing, a job I 
did every single day for a living? Perhaps they've invented this 
omnipotent Linux supercomputer from yesterday to today, that I just 
*somehow* am missing in my server rooms?

mlw wrote:

> "open sourced?" People have been licensing the source to UNIX for a long
> time.

I know! But you try explaining that to the Linux crowd. They're little slow.

> Linux, written by engineers for engineers.

Don't try to be funny, it's an insult to professional engineers.

Ku Karlovsky wrote:

> So you're only a UNIX admin after your job as a mechanic.

Actually I make full use of my brain, unlike some people (yes, I'm 
referring to you), so that in my spare time, I am able to work on my 
cars instead of paying somebody else because I'm too dumb/too lazy like 
most people.

Arschloch.

JEDIDIAH wrote:
> 	Changing history. How Orwellian of you.

And wouldn't you know it, George and I get along pretty damn well.

> 	The SGI Altix will scale to 1024 cpus running Linux and a single
> NUMA image of the operating system.

Forgot IRIX64, did you?

> 	What Sun/Fujistu hardware does this for Solaris?

StarFire? Computers/Informatics and you aren't PER TU, are you?

> 	...just another fanboy.

Actually, I'm not. I love SGI and IRIX. But the way Sun's been going 
lately, I just might become one.

> 	This little comment here only demonstrates that you are some sort
> of demented wannabe, not even a genuine (as a matter of experience) fanboy.

Show me.

On Wednesday 15 June 2005 10:26, Rich Teer <rich.teer@rite-group.com>
(<Pine.SOL.4.58.0506150926180.24062@zen.rite-group.com>) wrote:

> What else is new?!  ;-)

Your cross-posted spam isn't new!!!1!  ;-)  Does Solaris have a
newsreader that disables cross-posting by default?  You're failing to
do it manually, over and over again.

Jim Richardson wrote:

> It'd be nice if you were to come up with some technical reasons for
> dissing Linux, rather than you vague innuendo and handwaving. 

Would it really? Well then -

How about lack of consistency?
How about incompatible libCs?
How about inconsistent software management subsystems, and million 
packaging formats?
How about using `info` because `man` is "too corporate"?
How about having to create a symbollic link in /etc/ MANUALLY in order 
to set a proper time zone, and having that documented exactly NOWHERE?
How about a LACK of COMPLETE, CONSISTENT, PROFESSIONALY WRITTEN man pages?
How about totally inconsistent ABI/API/command interfaces and switches 
(and the Linux kernel developers practially *guaranteeing* it will stay 
that way)?
How about BLATANTLY DISREGARDING EVERY POSIX and SVR4 mandate?
How about PROPRIETARY management tools like YaST breaking production 
systems just by "updating" it (fails to handle config files properly).
How about using the name limiting /dev/sd /dev/hd convention (trying to 
imitate BSD and failing at it miserably)
How about the entire threading subsystem being ripped out, replaced and 
rendering the existing code and applications unusable without recompiling?
How about a million and one and a half forks of "Linux"?
How about... the list could go on and on. Amateurs!

Daniel Tryba wrote:

> Knoppix is doing a very good job at this.

"Knoppix" also happens to be a toy designed for promoting Linux.

> So you have never seen Debian/stable.

I believe he was describing Debian. It is the only Linux that makes an 
attempt to be stable/consistent, but it's impossible as long as the 
kernel developers are doing what they're doing. And they'll keep doing 
it, too, becuase they're no engineers; they're just hobbysts pretending 
to be developers.

Logan Shaw wrote:

> Then why does www.knoppix.org, the first hit when I type "knoppix"
> in on google.com, say that "KNOPPIX is a bootable CD" on a web
> page whose title is "KNOPPIX Linux Live CD"?  If it's more than a
> Live CD distribution, they are doing a good job hiding it.  Nothing

It used to be, they've added the install option in the recent releases.

The really embarrasing part of the Knoppix story is that the FreeBSD 
folks cut them to it by releasing "Freesbie", which was also a FreeBSD 
live CD but offered you the possibility of installing.
What's even funnier is that the BSD folks actually got into the game 
quite late. Well, that's why BSD is always going to be lightyears ahead 
from Linux. They've got some folks there with good heads on their shoulders.

Jim Richardson wrote:

> You can do much the same thing with Kanotix, Mepis, and PCLinuxOS to
> name but a few. For more info on livecds, of many flavours and uses,
> check out <http://www.frozentech.com/content/livecd.php>. There are
> distros for computer forensics, multimedia, rescue, small footprint
> systems, security, games, scientific analysis. &tc. 

You mean to tell him that there are now a zillion versions of live Linux 
CDs in addition to already existing trillion versions and forks of Linux 
distros?

Jim Richardson wrote:

> Actually, yes you can. In fact, this is what most of the distro's
> install image does. Using a combination of a well thought out initrd
> and a modular kernel. For example, I am using the same kernel/install
> image, on three different machines, with different chipsets, video, and
> ancillary componants. (Ubuntu Hoary in this case) 

What really interests me: since I have free and open Solaris, what is 
the advantage of Linux now?

> That's what the distros do. While there are certainly mistakes made, the
> distros vet the kernel (and other software) for stability, and security.
> If you want to roll your own, then you are on your own, which is how it
> should be. 

You should not have to roll out your own, that's the crux of the problem!

Marc wrote:

> PS: I am not trying to say linux is better, I prefer solaris, I just want
> people to chose it for right reasons.

OK, and according to you, which reasons would those be?

On Wed, 15 Jun 2005 21:06:27 +0200, UNIX admin wrote:

> Jim Richardson wrote:
> 
>> It'd be nice if you were to come up with some technical reasons for
>> dissing Linux, rather than you vague innuendo and handwaving. 
> 
> Would it really? Well then -
> 
> How about lack of consistency?

Consistency in what?

> How about incompatible libCs?
> How about inconsistent software management subsystems, and million 
> packaging formats?

Millions? I think not. There's about four or five basic formats that I'm
aware of. Most distros adhere to one or other of them.

> How about using `info` because `man` is "too corporate"?

??? Both are available, if required. I never heard of any of this 'too
corporate stuff'.

> How about having to create a symbollic link in /etc/ MANUALLY in order 
> to set a proper time zone, and having that documented exactly NOWHERE?

??? Never heard of that. I just click on 'Configure Timezone'.

> How about a LACK of COMPLETE, CONSISTENT, PROFESSIONALY WRITTEN man pages?

You're free to vulunteer to write them, if it's so important to you.

<snip>

> How about PROPRIETARY management tools like YaST breaking production 
> systems just by "updating" it (fails to handle config files properly).

Yast is no longer proprietary. No doubt that means it'll be fixed pretty
soon.

<snip>

> How about a million and one and a half forks of "Linux"?

Again, you exaggerate. I don't believe the kernel is forked at all. X has
forked, to the great improvement of it. Other software projects that run
on Linux may have forked, but I don't see what the great problem is. It's
not intended as one monolithic system

> How about... the list could go on and on. Amateurs!

Why do you assume the Linux developers are amateurs? Their goals may not
suit you, but that's life. They're not doing what they're doing to suit
you, but to suit themselves and their user base.

-- 
Kier

Ku Karlovsky wrote:

[snip]

Kid, do your mommy and daddy know that you're on the internet, instead 
of studying for tomorrow's highschool class exams?

If your daddy finds out, he might whip out some of that "Karlovsky 
Ku-Fu" on ya...

Kevin Bowling wrote:

>   HPUX, AIX, et al could run games or anything else for the matter if
> you port them.  Your proof is superfluous.  Market support will never
> rally behind Solaris for it to happen.

We will see.

On Wed, 15 Jun 2005 21:17:30 +0200, UNIX admin wrote:

> Jim Richardson wrote:
> 
>> You can do much the same thing with Kanotix, Mepis, and PCLinuxOS to
>> name but a few. For more info on livecds, of many flavours and uses,
>> check out <http://www.frozentech.com/content/livecd.php>. There are
>> distros for computer forensics, multimedia, rescue, small footprint
>> systems, security, games, scientific analysis. &tc. 
> 
> You mean to tell him that there are now a zillion versions of live Linux 
> CDs in addition to already existing trillion versions and forks of Linux 
> distros?

Why do you assume this is a bad thing? Contrary to popular myth, one size
does *not* fit all. Not that there are zillions or trillions, of course. A
couple of hundred or so, at most, I should imagine, all catering to
varying needs.

-- 
Kier

On Wed, 15 Jun 2005 20:11:59 +0200, UNIX admin
<tripivceta@hotmail.com> wrote in message
<<42b06f4e$0$1157$5402220f@news.sunrise.ch>>:

> Hello! It's not about what I *need* (I don't *need* AIX, I have 
> IRIX/IRIX64, HP-UX and Solaris for that).

You said you use Outlook Express at work.  You're an MCSE playing
"UNIX admin".

On Wed, 15 Jun 2005 21:06:27 +0200, UNIX admin
<tripivceta@hotmail.com> wrote in message
<<42b07c12$0$1150$5402220f@news.sunrise.ch>>:

> Would it really? Well then -

Stop spamming.  Your endless repetition of fairy tails and marketing
drivel is not useful to you or anyone else.

> How about... the list could go on and on. Amateurs!

Outlook Express user!

On Wed, 15 Jun 2005 21:13:42 +0200, UNIX admin wrote:

> Logan Shaw wrote:
> 
>> Then why does www.knoppix.org, the first hit when I type "knoppix"
>> in on google.com, say that "KNOPPIX is a bootable CD" on a web
>> page whose title is "KNOPPIX Linux Live CD"?  If it's more than a
>> Live CD distribution, they are doing a good job hiding it.  Nothing
> 
> It used to be, they've added the install option in the recent releases.

Since 3.3 or .4, I think, though someone will no doubt speak up if I'm
wrong. We're now on 3.9.

> 
> The really embarrasing part of the Knoppix story is that the FreeBSD 
> folks cut them to it by releasing "Freesbie", which was also a FreeBSD 
> live CD but offered you the possibility of installing.

I find it doubtful that Knoppix were beaten to it - Knoppix has been
installable for some time. I only just heard of FreeSBSie, and when I
tried it, it wasn't half as good as Knoppix at detecting my hardware. I
wouldn't trust it to install.

> What's even funnier is that the BSD folks actually got into the game 
> quite late. Well, that's why BSD is always going to be lightyears ahead 
> from Linux. They've got some folks there with good heads on their shoulders.

You're talking out of your arse.

-- 
Kier

On Wed, 15 Jun 2005 21:09:47 +0200, UNIX admin
<tripivceta@hotmail.com> wrote in message
<<42b07cd9$0$1150$5402220f@news.sunrise.ch>>:

> kernel developers are doing what they're doing. And they'll keep doing 
> it, too, becuase they're no engineers;

A purported "UNIX admin" is in no position to know who is an engineer
and who isn't.  You're low-level technical support.

> they're just hobbysts pretending to be developers.

You're just an Outlook Express-using MCSE pretending to be a "UNIX
admin".  Get over your petty self.

On Wed, 15 Jun 2005 21:13:42 +0200, UNIX admin
<tripivceta@hotmail.com> wrote in message
<<42b07dc5$0$1150$5402220f@news.sunrise.ch>>:

> The really embarrasing part 

is your pretense and spamming.

On Wed, 15 Jun 2005 21:20:58 +0200, UNIX admin
<tripivceta@hotmail.com> wrote in message
<<42b07f79$0$1162$5402220f@news.sunrise.ch>>:

> What really interests me: since I have free and open Solaris

You should take keep that interest where it belongs: in your
imagination.

> You should not have to roll out your own, that's the crux of the problem!

The crux of the problem is your cult-like ravings.

In comp.os.linux.advocacy UNIX admin <tripivceta@hotmail.com> wrote:
>> Knoppix is doing a very good job at this.
> 
> "Knoppix" also happens to be a toy designed for promoting Linux.

You may think of as a toy, for me it's a full featured rescue system.

>> So you have never seen Debian/stable.
> 
> I believe he was describing Debian. It is the only Linux that makes an 
> attempt to be stable/consistent, but it's impossible as long as the 
> kernel developers are doing what they're doing. And they'll keep doing 
> it, too, becuase they're no engineers; they're just hobbysts pretending 
> to be developers.

Please elaborate. As long as you are using kernels supplied by the
distribution what is the problem? They test it and integrate it in their
current version. 

On Wed, 15 Jun 2005 21:09:47 +0200, UNIX admin wrote:

> Daniel Tryba wrote:
> 
>> Knoppix is doing a very good job at this.
> 
> "Knoppix" also happens to be a toy designed for promoting Linux.

Perhaps you'd better tell Klaus Knopper that - he seems to be under teh
impression it's a fully-functional Linux distribution on a CD, something
you can carry everywhere ands still have your Linux desktop with you when
you don't have your Linux PC with you.

I've used Knoppix and several of its derivatives, and other Linux live
CDs. They certainly aren't toys. They *are* a good way to demonstrate the
strengths of Linux - it can be pretty stunning to someone who's never seen
Linux, to realise a whole OS can run productively off a single CD.

> 
>> So you have never seen Debian/stable.
> 
> I believe he was describing Debian. It is the only Linux that makes an 
> attempt to be stable/consistent, but it's impossible as long as the 
> kernel developers are doing what they're doing. And they'll keep doing 
> it, too, becuase they're no engineers; they're just hobbysts pretending 
> to be developers.

You seem woefully ill-informed. I'm sure all the developers of Linux would
love to know they're just 'hobbyists'. Especially the ones working for
Novell/SUSE, RedHat and Mandriva.

-- 
Kier 

Ku Karlovsky <nospam@nospam.nospam.not> writes:

> On Wed, 15 Jun 2005 20:11:59 +0200, UNIX admin
> <tripivceta@hotmail.com> wrote in message
> <<42b06f4e$0$1157$5402220f@news.sunrise.ch>>:
>
>> Hello! It's not about what I *need* (I don't *need* AIX, I have 
>> IRIX/IRIX64, HP-UX and Solaris for that).
>
> You said you use Outlook Express at work.  You're an MCSE playing
> "UNIX admin".

I am confused here: according to User-Agent field, UNIX Admin uses
"Mozilla/5.0 (X11; U; SunOS i86pc; en-US; rv:1.4) Gecko/20040414",
i.e. Solaris x86, whereas you, K.K. use "Forte Agent 1.92/32.572", a
product according to ForteInc web pages "Agent supports Windows 95, 98, ME,
NT, 2000, & XP" so you are using MS Windows.

Kettle calling pot black?

Bye, Dragan

-- 
Dragan Cvetkovic, 

To be or not to be is true. G. Boole      No it isn't.  L. E. J. Brouwer

!!! Sender/From address is bogus. Use reply-to one !!!

In article <lmbr6731vy.fsf@privacy.net>,
	Dragan Cvetkovic <me@privacy.net> writes:
> 
> ***************************************************************************
> * UNIX 03 CERTIFICATION ON 3 SIMULTANEOUS ARCHITECTURES
> ****************************************************************************
> 
> We are delighted to announce that Sun Microsystems has registered
> Solaris 10 on their SPARC and X86 platforms as conforming to the
> UNIX 03 Product Standard.
> 
> -- Solaris 10 Operating System - on 32-bit & 64-bit SPARC based systems

Hum, I wonder which 32-bit SPARC based systems those would be?
;-)

> -- Solaris 10 Operating System - on 32-bit X86 based system
> -- Solaris 10 Operating System - on 64-bit X86 based system

-- 
Andrew Gabriel

On 2005-06-15, Rich Teer <rich.teer@rite-group.com> wrote:
> On Wed, 15 Jun 2005, JEDIDIAH wrote:
>
>> 	Make up your damn mind. Either you want it to be a stable
>> configuration or mangled mess living on the bleeding edge.
>
> It's perfectly reasonable to want both: stable for production
> servers, and bleeding edge for one's own desktop.

	This is the value of genuine capitalism instead of a command
economy masquerading as capitalism.

-- 
	The best OS in the world is ultimately useless         |||
	if it is controlled by a Tramiel, Jobs or Gates.      / | \
                                                     

andrew@cucumber.demon.co.uk (Andrew Gabriel) writes:

> In article <lmbr6731vy.fsf@privacy.net>,
> 	Dragan Cvetkovic <me@privacy.net> writes:
>> 
>> ***************************************************************************
>> * UNIX 03 CERTIFICATION ON 3 SIMULTANEOUS ARCHITECTURES
>> ****************************************************************************
>> 
>> We are delighted to announce that Sun Microsystems has registered
>> Solaris 10 on their SPARC and X86 platforms as conforming to the
>> UNIX 03 Product Standard.
>> 
>> -- Solaris 10 Operating System - on 32-bit & 64-bit SPARC based systems
>
> Hum, I wonder which 32-bit SPARC based systems those would be?
> ;-)

Good eye. Interestingly enough,
e.g.
http://www.opengroup.org/csq/view.mhtml?norationale=1&noreferences=1&RID=sun%2FCX1%2F2
gives some details:

32-bit Intel platform: Dell Precision 220/Pentium II Operating Environment:
SunOS 5.10 

32-bit SPARC platform: Sun Fire V120 Operating Environment: SunOS 5.10

64-bit SPARC platform: Sun Ultra-5 Operating Environment: SunOS 5.10

For a lot more details, start at http://www.opengroup.org/csq/

Bye, Dragan

-- 
Dragan Cvetkovic, 

To be or not to be is true. G. Boole      No it isn't.  L. E. J. Brouwer

!!! Sender/From address is bogus. Use reply-to one !!!

On 2005-06-15, UNIX admin <tripivceta@hotmail.com> wrote:
> Martin Etteldorf wrote:
>> Maybe you should stay with your 32 bit machine instead if it is
>> sufficient for your needs.
>
> Hello! It's not about what I *need* (I don't *need* AIX, I have 
> IRIX/IRIX64, HP-UX and Solaris for that). It's about what I *want*, and 
> I *want* to learn AIX much deeper, but I'm not prepared to pay with my 
> liver and kidneys just to be able to get an aging 32-bit pSeries rack 
> mountable server.

	More clueless misinformation from McNealy's fanboy.

>
>> A B50 for $3000? You're joking. They're selling for less than
>> 500$ (or 400 Euros, for that matter) these days on ebay.
>
> Did you happen to look on ebay for a B50? Interestingly enough, good 
> luck finding one.
>
	
	I found a number of pseries machines actually.
	

-- 
	The best OS in the world is ultimately useless         |||
	if it is controlled by a Tramiel, Jobs or Gates.      / | \
                                                     

On 2005-06-15, UNIX admin <tripivceta@hotmail.com> wrote:
> JEDIDIAH wrote:
>   	
>> 	...been "officially supported" in the past. Many of us still
>> remember those days.
>
> Nooo, Solaris10 on x86 is officially supported by Oracle for Oracle10g.
>
> Hello. Check it.

 	...been "officially supported" in the past. Many of us still
	remember those days.

>
>> 	Larry doesn't merely "certify" Linux. It is his flagship high 
>> performance platform.
>
> For right now. But we will see where Larry ends up.

	Considering the obvious interest in integrated clusterware, IBM
has more of a chance than gaining Larry's attention in this regard.

>
>> 	What you said above doesn't have any relevance.
>> 
>> 	Linux is Larry's favorite son now. No amount of spin control will
>> change that.
>
> NOW. But Larry won't guarantee you it stays that way. If I know Larry, 
> and I know him, he'll ditch Linux as soon as he smells blood.

	Wake us when the largest Oracle clusters are being deployed
on Solaris again, not that we saw too many of those during the "reign
of Sun".

>
>> 	This is a bald faced lie on both counts?
>> 
>> 	Have you ever even touched a fireplane system? I doubt it.
>
> Do tell? Perhaps I'll learn something new about supercomputing, a job I 
> did every single day for a living? Perhaps they've invented this 
> omnipotent Linux supercomputer from yesterday to today, that I just 
> *somehow* am missing in my server rooms?

	Your blinders have no relevance to reality.

-- 
	The best OS in the world is ultimately useless         |||
	if it is controlled by a Tramiel, Jobs or Gates.      / | \
                                                     

On 2005-06-15, UNIX admin <tripivceta@hotmail.com> wrote:
> JEDIDIAH wrote:
>> 	Changing history. How Orwellian of you.
>
> And wouldn't you know it, George and I get along pretty damn well.
>
>> 	The SGI Altix will scale to 1024 cpus running Linux and a single
>> NUMA image of the operating system.
>
> Forgot IRIX64, did you?

	An Itanium version of Irix.

	Must have missed that one.

>
>> 	What Sun/Fujistu hardware does this for Solaris?
>
> StarFire? Computers/Informatics and you aren't PER TU, are you?

	Not quite.

>
>> 	...just another fanboy.
>
> Actually, I'm not. I love SGI and IRIX. But the way Sun's been going 
> lately, I just might become one.
>
>> 	This little comment here only demonstrates that you are some sort
>> of demented wannabe, not even a genuine (as a matter of experience) fanboy.
>
> Show me.

	SGI Altix 3000 vs. ???

-- 
	The best OS in the world is ultimately useless         |||
	if it is controlled by a Tramiel, Jobs or Gates.      / | \
                                                     

JEDIDIAH <jedi@nomad.mishnet> writes:

> On 2005-06-15, UNIX admin <tripivceta@hotmail.com> wrote:
>>
>> Nooo, Solaris10 on x86 is officially supported by Oracle for Oracle10g.
>>
>  	...been "officially supported" in the past. Many of us still
> 	remember those days.

From http://www.oracle.com/technology/support/metalink/index.html

General Notes For Oracle Database - Enterprise Edition On Solaris Operating
System x86: 

      Sun Solaris OS 10 Information:
          o The following database products are certified on the Sun
          Solaris 10 operating system: 
                + Oracle Database 9iR2 (with 9.2.0.4 and 9.2.0.6 Patchsets)
          for Sun Solaris SPARC32 
                + Oracle Database 9iR2 (with 9.2.0.5 and 9.2.0.6 Patchsets)
          for Sun Solaris SPARC64 
                + Oracle Database 10gR1 (with 10.1.0.3 Patchset) for Sun
          Solaris SPARC64 
                + Oracle Database 10gR1 (with 10.1.0.3 Patchset) for Sun
          Solaris x86 
          o When Oracle Database 10gR2 releases on Sun platforms later in
          2005, the following releases will be certified on the Sun Solaris
          10 operating system: 
                + 10gR2 Database for Sun Solaris SPARC64
                + 10gR2 Database for Sun Solaris AMD64
                + 10gR2 Database for Sun Solaris x86 

HTH, Dragan

-- 
Dragan Cvetkovic, 

To be or not to be is true. G. Boole      No it isn't.  L. E. J. Brouwer

!!! Sender/From address is bogus. Use reply-to one !!!

On 2005-06-15, UNIX admin <tripivceta@hotmail.com> wrote:
> Jim Richardson wrote:
>
>> Actually, yes you can. In fact, this is what most of the distro's
>> install image does. Using a combination of a well thought out initrd
>> and a modular kernel. For example, I am using the same kernel/install
>> image, on three different machines, with different chipsets, video, and
>> ancillary componants. (Ubuntu Hoary in this case) 
>
> What really interests me: since I have free and open Solaris, what is 
> the advantage of Linux now?

	Someone besides Sun to support it, including proprietary vendors.

[deletia]


-- 
	The best OS in the world is ultimately useless         |||
	if it is controlled by a Tramiel, Jobs or Gates.      / | \
                                                     

JEDIDIAH <jedi@nomad.mishnet> writes:

> On 2005-06-15, UNIX admin <tripivceta@hotmail.com> wrote:
>>
>> What really interests me: since I have free and open Solaris, what is 
>> the advantage of Linux now?
>
> 	Someone besides Sun to support it, including proprietary vendors.

LOL. Please read more carefully what you wrote.

Dragan

-- 
Dragan Cvetkovic, 

To be or not to be is true. G. Boole      No it isn't.  L. E. J. Brouwer

!!! Sender/From address is bogus. Use reply-to one !!!

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Wed, 15 Jun 2005 14:53:55 GMT,
 Rich Teer <rich.teer@rite-group.com> wrote:
> On Wed, 15 Jun 2005, JEDIDIAH wrote:
>
>> 	Make up your damn mind. Either you want it to be a stable
>> configuration or mangled mess living on the bleeding edge.
>
> It's perfectly reasonable to want both: stable for production
> servers, and bleeding edge for one's own desktop.
>


And several distros offer just that. Debian being one of them. 

However, the poster's qustion was regarding stability, which was
answered. It's odd that someone who has familiarity with with linux,
thinks that Debian stable doesn't fit that bill perfectly. 

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQFCsI+Gd90bcYOAWPYRAkc9AKC0JnK6qccyD1W5Ai65D2ZiQcAE8wCfWXA0
iKdZQrJQEK1Blrg15a04Cik=
=b/Ci
-----END PGP SIGNATURE-----

-- 
Jim Richardson     http://www.eskimo.com/~warlock
A positive attitude may not solve all your problems,
but it will annoy enough people to make it worth the effort.

UNIX admin wrote:
> Logan Shaw wrote:
>
> > Then why does www.knoppix.org, the first hit when I type "knoppix"
> > in on google.com, say that "KNOPPIX is a bootable CD" on a web
> > page whose title is "KNOPPIX Linux Live CD"?  If it's more than a
> > Live CD distribution, they are doing a good job hiding it.  Nothing
>
> It used to be, they've added the install option in the recent releases.

Wrong. Here is a wayback machine page from February 2003 mentioning
Knoppix hard disk install:
<URL:http://web.archive.org/web/20030209062904/http://www.knoppix.net/>.
'Knoppix makes a great GUI installer for Debian'
and
' knx-hdinstall 0.38 released'

Why did you lie when you could have checked the facts? Very trollish
behavior IMO.

> The really embarrasing part of the Knoppix story is that the FreeBSD
> folks cut them to it by releasing "Freesbie", which was also a FreeBSD
> live CD but offered you the possibility of installing.

Wrong again. From FreeSBIE FAQ:
'Since FreeSBIE 1.1 it's possible to use the BSDInstaller to install
FreeSBIE on your hard drive, and then turn it into FreeBSD 5.3-STABLE
by means of cvsup'

When was FreeSBIE 1.1 released?
<URL:http://www.freesbie.org/doc/1.1/ANNOUNCE.txt> is from December 6th
2004, almost two years later than the Knoppix hard disk installer page.
So you lied again.

Now why would you lie twice in the same message? Did you think no one
would check? That would be idiotic. Any *real* UNIX admin would have
checked the facts first. And these were very easy to find facts. Could
it be that you are not a real UNIX admin? Yes. What would you be if you
were not a real UNIX admin? A troll.

-- 
Timo Pirinen
pirisisi@dlc.fi

Kier wrote:
> On Wed, 15 Jun 2005 21:09:47 +0200, UNIX admin wrote:

>>I believe he was describing Debian. It is the only Linux that makes an 
>>attempt to be stable/consistent, but it's impossible as long as the 
>>kernel developers are doing what they're doing. And they'll keep doing 
>>it, too, becuase they're no engineers; they're just hobbysts pretending 
>>to be developers.
> 
> 
> You seem woefully ill-informed. I'm sure all the developers of Linux would
> love to know they're just 'hobbyists'. Especially the ones working for
> Novell/SUSE, RedHat and Mandriva.

Not to mention the ones working for IBM and HP.

- Greg
-- 
Greg Cagle
gregc at gregcagle dot com

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Wed, 15 Jun 2005 21:06:27 +0200,
 UNIX admin <tripivceta@hotmail.com> wrote:
> Jim Richardson wrote:
>
>> It'd be nice if you were to come up with some technical reasons for
>> dissing Linux, rather than you vague innuendo and handwaving. 
>
> Would it really? Well then -
>
> How about lack of consistency?

in Linux? sounds like you have a gripe with some distros, not Linux,
pick a distro, there, consistant. 

> How about incompatible libCs?

see above

> How about inconsistent software management subsystems, and million 
> packaging formats?

see above

> How about using `info` because `man` is "too corporate"?

I use man, on the rare occasions where info is all that is available,
(and I agree that info sucks in many ways) pinfo is my reader of choice,
when I am not using the GUI help system in GNOME or KDE which integrates
them. 

> How about having to create a symbollic link in /etc/ MANUALLY in order 
> to set a proper time zone, and having that documented exactly NOWHERE?

Hmm, I rt click on the clock on the app bar, select preferences, and
click on the timezone. 

> How about a LACK of COMPLETE, CONSISTENT, PROFESSIONALY WRITTEN man pages?

Are the man pages for gnu utils better on say, solaris?

> How about totally inconsistent ABI/API/command interfaces and switches 
> (and the Linux kernel developers practially *guaranteeing* it will stay 
> that way)?

there are good reasons to not support stable *kernel* level abi's,
they've been discussed at length, many times.


> How about BLATANTLY DISREGARDING EVERY POSIX and SVR4 mandate?

???

> How about PROPRIETARY management tools like YaST breaking production 
> systems just by "updating" it (fails to handle config files properly).

yast is open source now, how is it proprietary? If it causes errors,
file bugfixes, are you going to claim that Solaris is 100% bug free? 

> How about using the name limiting /dev/sd /dev/hd convention (trying to 
> imitate BSD and failing at it miserably)

not sure what you mean here either. 

> How about the entire threading subsystem being ripped out, replaced and 
> rendering the existing code and applications unusable without recompiling?

major changes to kernel level code require's a recompile, when the
kernel changes, If you prefer, keep the old kernel, and use backports. 

> How about a million and one and a half forks of "Linux"?

? do you mean that you are complaining that there are distro choices? 

> How about... the list could go on and on. Amateurs!

Might want to give that caps lock key a break for a while. 




-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQFCsJHRd90bcYOAWPYRAnpcAJ42dP7vqsO84uwuWHUHmOX0H4qqzACfTmFb
grZQ+W9jw5lr2Hckd1MyLC0=
=Vamo
-----END PGP SIGNATURE-----

-- 
Jim Richardson     http://www.eskimo.com/~warlock
"The dinosaurs became extinct because they didn't have a space program"
	-- Larry Niven

begin  virus.txt.scr UNIX admin wrote:

> Kevin Bowling wrote:
> 
>>   HPUX, AIX, et al could run games or anything else for the matter if
>> you port them.  Your proof is superfluous.  Market support will never
>> rally behind Solaris for it to happen.
> 
> We will see.

Right. Until then, crawl back under that stone you came from

-- 
No trees were destroyed in the sending of this message, however, a
significant number of electrons were terribly inconvenienced.

begin  virus.txt.scr Kier wrote:

> On Wed, 15 Jun 2005 21:06:27 +0200, UNIX admin wrote:
< snip >
 
>> How about PROPRIETARY management tools like YaST breaking production
>> systems just by "updating" it (fails to handle config files properly).
> 
> Yast is no longer proprietary. No doubt that means it'll be fixed pretty
> soon.
> 

It is fixed, and since quite some time
If it finds a config not changed by itself, it will write the new one with a
suffix and leave the other alone. From then on, you will be reminded during
boot that there are still "unresolved" config changes, and where

This "Unix admin" should stay to his precious Slowaris and stop talking
nonsense

< snip more ridiculous bullshir > 

-- 
Only two things are infinite,
 the Universe and Stupidity.
And I'm not quite sure about the former.
        - Albert Einstein

On Wed, 15 Jun 2005, Casper H.S. Dik wrote:

> We're planning to release most of it as part of OpenSOlaris; perhaps
> I can check today whether I can release the stuff I personnally wrote

Great!

> under the CDDL rightaway.  (It would be prototype stuff, but it
> is kinda useful to have power management and battery monitoring on
> your Ferrari)

Treu, altrhough for me personally, I find the Wifi stuff more useful.

-- 
Rich Teer, SCNA, SCSA, OpenSolaris CAB member

President,
Rite Online Inc.

Voice: +1 (250) 979-1638
URL: http://www.rite-group.com/rich

UNIX admin <tripivceta@hotmail.com> writes:

>Ku Karlovsky wrote:

>> OpenSolaris IS NOT UNIX BUT YOU DON'T UNDERSTAND.  The OpenGroup
>> doesn't list Solaris 10 or any of its marketing derivates as
>> qualified.

>Looky here, OpenSolaris is a 1:1 exact copy of Solaris, which is UNIX 
>certified. Solaris10 may not be explicitly certified, but since, unlike 
>crappy Linux, Sun makes sure latest Solaris is backward compatible, we 
>do in fact run certified UNIX.

Solaris 10 is listed:

http://www.opengroup.org/openbrand/register/


Casper
-- 
Expressed in this posting are my opinions.  They are in no way related
to opinions held by my employer, Sun Microsystems.
Statements on Sun products included here are not gospel and may
be fiction rather than truth.

Dragan Cvetkovic <me@privacy.net> wrote:
> 32-bit SPARC platform: Sun Fire V120 Operating Environment: SunOS 5.10

Interesting. I thought S10 got rid of 32bit kernel on SPARC...
-- 
#pragma ident "Seongbae Park, compiler, http://blogs.sun.com/seongbae/"

In article <d8q97k$o24$2@news1nwk.SFbay.Sun.COM>,
Seongbae Park  <Seongbae.Park@Sun.COM> wrote:
>Dragan Cvetkovic <me@privacy.net> wrote:
>> 32-bit SPARC platform: Sun Fire V120 Operating Environment: SunOS 5.10
>
>Interesting. I thought S10 got rid of 32bit kernel on SPARC...

POSIX is a _source_ standard, so maybe they did just check a 32 bit
userland.

-- 
EMail:joerg@schily.isdn.cs.tu-berlin.de (home) J�rg Schilling D-13353 Berlin
      js@cs.tu-berlin.de		(uni)  
      schilling@fokus.fraunhofer.de	(work) Blog: http://schily.blogspot.com/
URL:  http://cdrecord.berlios.de/old/private/ ftp://ftp.berlios.de/pub/schily

In comp.os.linux.advocacy, Daniel Tryba
<partmapsswen@invalid.tryba.nl>
 wrote
on 15 Jun 2005 19:42:27 GMT
<42b084a3$0$10403$c5fe704e@news6.xs4all.nl>:
> In comp.os.linux.advocacy UNIX admin <tripivceta@hotmail.com> wrote:
>>> Knoppix is doing a very good job at this.
>> 
>> "Knoppix" also happens to be a toy designed for promoting Linux.
>
> You may think of as a toy, for me it's a full featured rescue system.

Well yes, of course it's a toy...it doesn't have the
official Microsoft Holographic Seal of Approval(r)(tm),
or the Unix Official Holographic Plaque(r)(tm).  Can't be
anything without the Microsoft Holographic Seal of
Approval(r)(tm), or the Unix Official Holographic
Plaque(r)(tm).

And if you act now we'll throw in a pinball game and search
doggy for free!

:-)

>
>>> So you have never seen Debian/stable.
>> 
>> I believe he was describing Debian. It is the only Linux that makes an 
>> attempt to be stable/consistent, but it's impossible as long as the 
>> kernel developers are doing what they're doing. And they'll keep doing 
>> it, too, becuase they're no engineers; they're just hobbysts pretending 
>> to be developers.
>
> Please elaborate. As long as you are using kernels supplied by the
> distribution what is the problem? They test it and integrate it in their
> current version. 
>

Color *me* confused, anyway.  There's only so many ways to write
an argument list to, say, fork(). :-)

Let's see now...hmm...uh...one...well...erm...yeah...one.

Exactly one.  Dare I count again? :-)

Granted, the engineers/hobbyists/gurus/superkernelhackers/interested
parties can probably screw up the implementation a dozen times over,
but that's a different issue, and I suspect that's unlikely anyway;
fork()'s not exactly one of the newer calls.

I think UNIX Admin's been looking at http://www.wehavethewayout.com
too long.  :-)

(Then again, so has everyone else, apparently; the website is
no longer available?  Not that http://www.wehavethewayin.com
is in much better shape -- a squatter got it.  Want a timeshare? :-) )

-- 
#191, ewill3@earthlink.net
It's still legal to go .sigless.

JEDIDIAH wrote:
> On 2005-06-15, Logan Shaw <lshaw-usenet@austin.rr.com> wrote:
> 
>>Daniel Tryba wrote:
>>
>>>In comp.os.linux.advocacy Logan Shaw <lshaw-usenet@austin.rr.com> wrote:
> 
> [deletia]
> 
>>>>When I checked out Debian and installed it as a test, it seemed
>>>>questionable whether it was a viable distribution anymore.
>>>
>>>
>>>Now what do you want? Stable is extremly stable, AFAIK the only upgraded
>>>package in 3.0 was ssh (updated after some panic attack from Theo (which
>>>proved unnecessary since that version wasn't vulnerable IIRC)) during
>>>it's 3 year lifecycle
> 
> 
> 	Make up your damn mind. Either you want it to be a stable
> configuration or mangled mess living on the bleeding edge.

Maybe I wasn't clear about what I wanted.  I want two things with
regards to releases:
(1)  For there to be a stable branch to which only bugfixes are
      being released.
(2)  Confidence that the software will continue to be developed
      and that periodically there will be new stable releases.
      (Even if that period is only every 2 years, that's OK.)

At the time I looked at Debian, it had #1, but it really wasn't
clear that it had #2.  People were talking about 3.0, but it
wasn't clear to me that it really would ever materialize.

Hmm, while I'm at it, I really want a third thing too:

(3)  Every stable branch to be supported for at least 3 or 4 years,
      i.e. for stable branches that do NOT automatically go
      unmaintained because they are not the *latest* stable branch.

I find that on point #3, OSes vary a lot.  Mac OS X, Solaris, AIX,
and Windows do a pretty good job with #3.  On OS X, you can still
get security updates and things for 10.2, even though they are
on 10.4 now.  With Solaris, you can still get updates to Solaris 8
and maybe even Solaris 7, even though they are on Solaris 10 now.
Microsoft is still updating Windows 98 a little bit.

Other operating systems have tended not to do as well for supporting
older stable branches.  SGI used to drop support for older releases
of IRIX with startling speed, which was made worse by the fact that
newer releases often couldn't run on older hardware.  (This meant that
when an SGI machine was about 4 years old, it was no longer viable for
much of anything.)

Overall, I don't have a positive impression of Linux in this category.
It seems that once a new stable branch is out, support for the
previous stable branch drops off rapidly.  Yes, the 2.4 kernel series
is still being developed, and that's good, but the experience I've
had with distributions is that you want your software to be supported
with updates for security and other serious bugs, you have to stay
on the upgrade treadmill.  This is fine for some machines, but in
many cases (especially servers) the wisest policy is to wait for
a stable branch to mature for 6 months or a year before moving to it.

Please note that if someone can point me to a Linux distribution
that does maintain stable branches for several years (at a minimum
3 years, I'd say), I'd like to know about it.  I'm not saying that
Linux isn't viable in this area.  I'm just saying that I went looking
for such a distribution and wasn't able to find a distribution that
clearly really does maintain stable branches for a long time.

   - Logan

Marc wrote:
> Logan Shaw wrote:

>>* You can't just build a universal kernel that contains modular
>>   support for everything in the source code, which is demand
>>   loaded as the hardware is found.  Instead, you have to customize
>>   the kernel for every machine or live with bloat.  This is not
>>   so wonderful if you have hundreds of machines, some of which
>>   are the same as each other and some of which are not.

> This is not true. I will speak only about the debian kernel because it is
> the one I know best, but it is more general. You can chose for most
> functionalities whether you want them in the kernel or as modules, and
> distributions usually put plenty of things in modules, and then have
> programs to load the apropriate modules for your machine.

I'm not sure why I would want programs to load the appropriate modules.
I'd really prefer that the kernel knew what the modules were for and
just loads them when it needs them (and unloads them when it no longer
needs them).  For example, if a module is a device driver, the kernel
should know the types of devices that the module supports without
loading the module (through a module config file, or through some
database, whatever, as long as it can be updated when a module is
added to the system).  Then when the kernel enumerates the buses and
finds the device IDs, it would know exactly which modules to load.
If the kernel module is a filesystem type, the kernel would load it
whenever it receives a request related to that filesystem type
(such as a request to mount a FAT filesystem).

Anyway, since it apparently is possible, I will change my complaint.
My complaint is that, even if it is possible, it's a giant pain in
the ass.  I spent several hours once going through all configuration
options I could find and making everything into a module that looked
like it could be a module.  I finally built a kernel configuration
file and used that to try to build a kernel.  I did not get a usable
kernel, because the build process actually failed before it even
finished building the kernel.

I concluded that even though it might be possible, it wouldn't be
practical.  Even though I had previous experience at configuring
kernels on BSD systems (on NetBSD and also the old SunOS 4.x
systems), doing the obvious things did not lead to a working
kernel.  Furthermore, since this was not the default configuration
by any means, I expect that if I ever did get it to work, it would
probably require more time investment to iron out all the kinks
than simply building a custom kernel and rebuilding the custom
kernel when things change and force me to do so.

So, my personal experience was that Linux forces you to choose
between two tedious manual processes:  learning all the intricacies
of building and using a fully-modular kernel (since that is not
the default and it requires non-trivial effort to make it happen),
or deal with making a custom kernel for each machine.

Along the way, I discovered some other silly things about maintaining
a kernel build for your system.  The first and most obvious is that
there is apparently no support for using the kernel configuration
name as a parameter for building, installing, and loading the kernel
itself and its modules.  Since the kernel is configurable and there
is a need to create different configurations based off the same kernel
version, this is a surprising oversight.  For example, to get decent
IDE performance, I found I needed to build a special kernel with
support for my motherboard's VIA chipset.  So, I already have a
production kernel, whose configuration I will call "vanilla".  I now
want to build another kernel based on a separate experimental configuration
I will call "with-via", and I want to test that kernel to see if it
works well to become the new production kernel.  In order to do this,
I need to have a mechanism to boot into the new kernel, and then if
I don't like it, I need to be able to revert to exact set of files
I was using before I did the experiment so that I have my production
kernel back.

But, because of the way the Makefiles on Linux work, if I do create
these two configurations, the build for both will take place in
/usr/src/linux-2.6.7.  When I build my new experimental configuration,
it will clobber all the files produced when I built my production
kernel.  And, worse, the modules for BOTH configurations will go
into /lib/modules/linux-2.6.7.  So, when I do my "make modules_install",
the experimental kernel's modules will overwrite the production
kernel's modules.  When I switch back to the production kernel,
I could easily be mixing and matching modules in a way that might
not be safe and certainly doesn't satisfy the aim of returning to
my previous configuration to ensure stability of the system.

Worse yet, I have to discover this by trial and error.  Sure, I
can learn this lesson, and if I'm lucky I won't have to learn it
the hard way while working on an important server.  But it
undermines my confidence in the system:  if I had to learn about
this issue with having two different configurations of the same
kernel sources installed, what other issues might be lurking there?
It's clear that the people who came up with the kernel build
process didn't think this out or didn't care about this issue
very much, so by trying to switch between configurations safely,
I am blazing a new trail and should expect there to be problems.

Even worse than that, the BSD kernel build system had this sorted
out over a decade ago.  When you create a BSD kernel configuration,
you put that configuration in a file and give that file a name.
Then you run the "config" command against that configuration file:
this creates a new directory hierarchy in which to build the object
files for the new configuration, keeping them totally separate from
the files used to build a different configuration of the same kernel
sources.  You can then take the resulting vmunix file built out of
that configuration tree and install it alongside an existing kernel
and switch between them at will, without one interfering with the
other.  This all happens easily out of the box.  BSD didn't have
kernel modules, and having modules makes the task slightly more
difficult, but since each configuration produces its own kernel
executable, it would be trivially easy to encode the configuration
name (or pathname to the configuration's modules dir) and sort
out the proper place to look for modules.  That this simple idea
didn't even appear to occur to the people who maintain the kernel
build system makes me wonder if the people who work on the kernel
have learned anything from history or ever had to be responsible
for the uptime of an important server.  If they have done neither,
that is IMHO a bad sign!

Along about here, someone is going to be tempted to suggest that
my problem can be solved by using $EXTRAVERSION.  I'd just like
to point out that I considered that, but it doesn't work well
for a variety of reasons.  One of the chief reasons is that
many Linux distros already use EXTRAVERSION for their own
modifications to the *source* of the kernel.  And that is what
it is intended to be:  something that marks the version of
the sources, not something that indicates which configuration
of the sources you have chosen to build.  Because of that, the
value of EXTRAVERSION shows up inside the output of "uname -r"
and all kind of other places that a kernel configuration name
shouldn't show up.  Using EXTRAVERSION for this purpose is at best
a hack to work around a limitation that Linux shouldn't have in
the first place.

Need I go on?  Another fun thing with Linux kernel modules and
the build system is that there is no stable binary kernel
interface.  This means all the modules you use on a system
should be built at the same time by the same compiler.  Even
though the modules can be demand-loaded, the system is really
not set up, so far as I can see, to allow you to build an
additional module at runtime and then load it into the kernel
at runtime, even though that would seem to be one of the
advantages of creating loadable kernel modules.  It also makes
it difficult for people to distribute binary kernel modules.
Yes, you can say that people should always be distributing
source (for philosophical reasons) and that therefore the
requirement to have the source of any kernel module is actually
a good thing.  But it's not true if you want to, say, use a
device driver from a third party that wants to support Linux
but doesn't want to release an open-source driver.  In the
real world, that's a silly limitation.  It forces people like
nVidia to create ad hoc hackish solutions where they add their
own layer of binary stability inside their modules, so that
they can distribute binaries of the real guts and source code
of the glue that maintains their own stable binary interface.
 From a technical point of view, that sucks because it causes
duplication of effort and bloated code (since everyone has to
reinvent this technology every time they want to have a
stable binary interface for something).  It also sucks because
it creates extra work for people who want to distribute binary
drivers and thus creates a disincentive for hardware companies
to support Linux if their business reality only allows them to
choose between closed-source drivers and no drivers.

   - Logan

On Wed, 15 Jun 2005 20:39:07 +0200, UNIX admin wrote:

> Liam Slider wrote:
> 
>> Ok, ok, but how large are those clusters? Also, the other part of my
>> original point still stands. No arm twisting, extorting, evil empire
>> keeping Linux dominant in super-computers.
> 
> Obviously you've never attended boardroom meetings where these types of 
> deals are made. But I have, and let me tell you, there's "a little more" 
> than just armtwisting tactics going on.

And we're to take your word for it? A known liar?

Arkady Duntov wrote:
> On Wednesday 15 June 2005 00:21, Logan Shaw <lshaw-usenet@austin.rr.com>
> (<hNPre.55866$6g3.4911@tornado.texas.rr.com>) wrote:

>>   Instead, you have to customize 
>>   the kernel for every machine or live with bloat.

> You don't have to build a customized kernel, but you can.

Yes, that's what I said, isn't it?  I said you have a choice:
to build a customized kernel and live with it wasting memory
for drivers and other bits that never will be used, OR to
build a custom kernel.

Meanwhile, on many other OSes, you don't need to do either.

>>* Users have to keep up to date on which release kernels are
>>   too buggy to run without crashing the machine.
> 
> 
> With the exception of 2.6.8 and its broken NFS support, you're wrong again. 
> (The bug was fixed in about twenty-four hours.)

No, I'm referring to 2.6.8 and its broken support for the
"keep running without hanging the system at random times" feature.
When I installed 2.6.8, I used substantially the same configuration
as from 2.6.7 (not identical, because the kernel versions need
slightly different configs), and with the 2.6.8 kernel, the system
kept hanging when I'd burn CDs and at other random times with the
2.6.8 kernel but not with the 2.6.7.  I was not using NFS at all.
I believe I tried again with 2.6.8.1, thinking it'd have bug fixes,
and I had the same problems.  The problems went away with the 2.6.10
kernel.

And, prior to that, people had similar issues with the 2.6 kernel
versus the 2.4 kernel.  The 2.6 series moved from 2.5 (development)
to 2.6 (stable), and even after it was declared stable, it was
still not really ready for prime time.  As a user, you had to know
that even though 2.6 said "stable", it really was better to stay
with 2.4 if you wanted something that was actually stable.  This
situation has now passed and 2.6 is ready for prime time (or at
least 2.6.7 was and 2.6.10 is) but the track record has not been
good, and you still cannot assume that just because something is
marked as "stable" means that it actually is.


By the way, thanks for being a pal and changing the followups-to
header to "comp.sys.mac.advocacy" and not making mention of that
in your post.  I've changed the newsgroups header back to include
comp.os.linux.advocacy again so that you can see this reply.

   - Logan

In article <0R2se.59259$6g3.54083@tornado.texas.rr.com>,
 Logan Shaw <lshaw-usenet@austin.rr.com> wrote:
> Hmm, while I'm at it, I really want a third thing too:
> 
> (3)  Every stable branch to be supported for at least 3 or 4 years,
>       i.e. for stable branches that do NOT automatically go
>       unmaintained because they are not the *latest* stable branch.
> 
> I find that on point #3, OSes vary a lot.  Mac OS X, Solaris, AIX,
> and Windows do a pretty good job with #3.  On OS X, you can still
> get security updates and things for 10.2, even though they are
> on 10.4 now.  With Solaris, you can still get updates to Solaris 8
> and maybe even Solaris 7, even though they are on Solaris 10 now.
> Microsoft is still updating Windows 98 a little bit.

You can still get updates to Redhat 7.3 from Fedora Legacy.

-- 
--Tim Smith

UNIX admin wrote:

> How about PROPRIETARY management tools like YaST 

You don't follow the news a lot, do you? You only read the comics,
right? 

YaST's license has been GPL for over a year. I'm sure any *real* UNIX
admin would have known that.

Novell Press Release from Marc 22, 2004:
<URL:http://www.novell.com/news/press/archive/2004/03/pr04026.html>
'... By contributing the industry's leading management tool, YaST, from
Novell's SUSE� LINUX business unit to the General Public License,
Novell will provide a platform that allows...'


-- 
Timo Pirinen
pirisisi@dlc.fi

Logan Shaw wrote:
> JEDIDIAH wrote:
> > On 2005-06-15, Logan Shaw <lshaw-usenet@austin.rr.com> wrote:
> >
> >>Daniel Tryba wrote:
> >>
> >>>In comp.os.linux.advocacy Logan Shaw <lshaw-usenet@austin.rr.com> wrote:
> >
> > [deletia]
> >
> >>>>When I checked out Debian and installed it as a test, it seemed
> >>>>questionable whether it was a viable distribution anymore.
> >>>
> >>>
> >>>Now what do you want? Stable is extremly stable, AFAIK the only upgraded
> >>>package in 3.0 was ssh (updated after some panic attack from Theo (which
> >>>proved unnecessary since that version wasn't vulnerable IIRC)) during
> >>>it's 3 year lifecycle
> >
> >
> > 	Make up your damn mind. Either you want it to be a stable
> > configuration or mangled mess living on the bleeding edge.
>
The problem isn't with new developments. Every OS brings in new
technologies -not just Linux alone. The problem is with an overhaul of
the entire architecture multiple times in a v short time frame. Its
kind of -"put in the latest greatest things and we should be fine".
What is the latest and greatest technology won't be the same ina few
yrs time. So, when architecting stuff., you need to do it in such a way
that future developments fit into the existing architecture or that the
stuff you are oputting in don\esn't overhaul the architecture
inconveniencing end users.

> Maybe I wasn't clear about what I wanted.  I want two things with
> regards to releases:
> (1)  For there to be a stable branch to which only bugfixes are
>       being released.
> (2)  Confidence that the software will continue to be developed
>       and that periodically there will be new stable releases.
>       (Even if that period is only every 2 years, that's OK.)
>
> At the time I looked at Debian, it had #1, but it really wasn't
> clear that it had #2.  People were talking about 3.0, but it
> wasn't clear to me that it really would ever materialize.
>
> Hmm, while I'm at it, I really want a third thing too:
>
> (3)  Every stable branch to be supported for at least 3 or 4 years,
>       i.e. for stable branches that do NOT automatically go
>       unmaintained because they are not the *latest* stable branch.
>
Try *bsd. Its a lot more stable and less crammed with goodies i.e. has
just the bare min code that could be required.


> I find that on point #3, OSes vary a lot.  Mac OS X, Solaris, AIX,
> and Windows do a pretty good job with #3.  On OS X, you can still
> get security updates and things for 10.2, even though they are
> on 10.4 now.  With Solaris, you can still get updates to Solaris 8
> and maybe even Solaris 7, even though they are on Solaris 10 now.
> Microsoft is still updating Windows 98 a little bit.
>
The OS with the least amt of security updates is OpenBSD -thats because
it has substantial portions of it written from the ground up to be
secure.
i.e if it is secure -it won't be hit with a DoS.

> Other operating systems have tended not to do as well for supporting
> older stable branches.  SGI used to drop support for older releases
> of IRIX with startling speed, which was made worse by the fact that
> newer releases often couldn't run on older hardware.  (This meant that
> when an SGI machine was about 4 years old, it was no longer viable for
> much of anything.)
>
> Overall, I don't have a positive impression of Linux in this category.
> It seems that once a new stable branch is out, support for the
> previous stable branch drops off rapidly.  Yes, the 2.4 kernel series

Partly the amateur mentality and partly because they aren't obliged to
support something they consider out-moded.

regards
-kamal

Kamal R. Prasad wrote:

> The problem isn't with new developments. Every OS brings in new
> technologies -not just Linux alone. The problem is with an overhaul of
> the entire architecture multiple times in a v short time frame. Its
> kind of -"put in the latest greatest things and we should be fine".
> What is the latest and greatest technology won't be the same ina few
> yrs time. So, when architecting stuff., you need to do it in such a way
> that future developments fit into the existing architecture or that the
> stuff you are oputting in don\esn't overhaul the architecture
> inconveniencing end users.

This is where the .NET and java virtual machines come in.

The OS, and its need to be 'stable',  is less critical for these type of 
application issues...so long as it can support the virtualization layer 
( java or .NET ).

It's kind of odd, that people /sell/ the OS, but *give* *away* .NET and 
java.   To me the later are far more valuable.

Virtualization comes with an overhead and language constraints -which
may not be acceptable to everyone.

 "It's kind of odd, that people /sell/ the OS, but *give* *away* .NET
and
java.   To me the later are far more valuable"

The complexity involved in getting the OS ported to a hw platform is a
lot more than getting a virtual machine running on TOP of an OS. So,
its not surprising as to why things are the way they are.

regards'
-kamal