Despite the subject line, I still think Solaris
is a great OS, but:

1. The famous troll Mike Cox is pretending to promote
Solaris when he's not pretending to promote Windows.
That can't be good.

2. Sun recently announced yet another 'business alliance'
with Microsoft.  There can't be a quicker way to achieve
bankruptcy than to team up with MS.

3. As a recent thread made so clear, thinking that opensourcing
Solaris is going to help Sun, is a crackheads delusion.

Sun needs to build faster, less costly Sparc hardware.
They need Sparc hardware that is competitive on the desktop
and in the server world.  They need to keep after Adobe,
Real Networks, and other proprietary vendors and make sure
Sparc machines don't become irrelevant.

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

> Despite the subject line, I still think Solaris
> is a great OS, but:
> 
> 1. The famous troll Mike Cox is pretending to promote
> Solaris when he's not pretending to promote Windows.
> That can't be good.
> 
> 2. Sun recently announced yet another 'business alliance'
> with Microsoft.  There can't be a quicker way to achieve
> bankruptcy than to team up with MS.
> 
> 3. As a recent thread made so clear, thinking that opensourcing
> Solaris is going to help Sun, is a crackheads delusion.
> 
> Sun needs to build faster, less costly Sparc hardware.
> They need Sparc hardware that is competitive on the desktop
> and in the server world.  They need to keep after Adobe,
> Real Networks, and other proprietary vendors and make sure
> Sparc machines don't become irrelevant.


Hell I'd be happy if they just ported their x86 drivers to SPARC.

Gregm

Dan Espen <daneNO@spam.mk.telcordia.com> wrote:
> 2. Sun recently announced yet another 'business alliance'
> with Microsoft.  There can't be a quicker way to achieve
> bankruptcy than to team up with MS.

Must make nice until the check clears :-)

-- 
Geoff Lane
Assume plicam damnatam, o tu moles muscaria muscerdarum -- Henry Beard

On Thu, 19 May 2005, Greg Menke wrote:

> Hell I'd be happy if they just ported their x86 drivers to SPARC.

And vice-versa...

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

President,
Rite Online Inc.

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

In article <ic8y2bfj0u.fsf@mk.telcordia.com>,
Dan Espen  <daneNO@SPAM.mk.telcordia.com> wrote:
>1. The famous troll Mike Cox is pretending to promote

If it's a kook, then you're feeding its psychosis by naming it in
plublic.

Please don't.

Thanks,
John
groenveld@acm.org

In article <428cd16e$0$38045$5a6aecb4@news.aaisp.net.uk>,
Geoff Lane  <zzassgl@buffy.sighup.org.uk> wrote:
>Must make nice until the check clears :-)

As far as the US taxman is concerned, the check cleared over a year ago.
<URL:http://www.theregister.co.uk/2004/07/20/sun_q4_04/>
<URL:http://www.sun.com/aboutsun/investor/annual_reports/sun_10k04.pdf>
| On April 1,2004,we entered into several agreements with
| Microsoft,including an agreement to settle all pending litigation
| between the two companies,a patent covenant and stand-still
| agreement,and a technical collaboration agreement.As further described
| in Note 13 to the Consolidated Financial Statements,we received $1,950
| million in cash and recognized approximately $1.6 billion in
| settlement income during the fourth quarter of fiscal 2004.

John
groenveld@acm.org

In article <ic8y2bfj0u.fsf@mk.telcordia.com>,
Dan Espen  <daneNO@SPAM.mk.telcordia.com> wrote:
>3. As a recent thread made so clear, thinking that opensourcing
>Solaris is going to help Sun, is a crackheads delusion.

Can you summarize or point to a good summary of that thread?

John
groenveld@acm.org

Dan Espen <daneNO@spam.mk.telcordia.com> wrote:
> 
> Despite the subject line, I still think Solaris
> is a great OS, but:
> 
> 1. The famous troll Mike Cox is pretending to promote
> Solaris when he's not pretending to promote Windows.
> That can't be good.

Mike is just a sad little troll. I'm afraid that the affairs of people
who can't even manage net-kook status aren't really a fair measure of
Sun.

> 2. Sun recently announced yet another 'business alliance'
> with Microsoft.  There can't be a quicker way to achieve
> bankruptcy than to team up with MS.

This one scares me too--especially that picture of Mcnealy and Ballmer
on Sun's homepage. Steve's expression looks very much like a ghoul who
has just eaten a particularly juicy human heart.

OK picture aside, I can't imagine why Sun can expect to survive a close
development partnership with Microsoft when no other company has managed
to do so in 30 years. You have to know that Microsoft wouldn't even bother
dealing closely with Sun unless they can either buy or crush them.

> 3. As a recent thread made so clear, thinking that opensourcing
> Solaris is going to help Sun, is a crackheads delusion.

I don't know about that. What I _do_ expect though, is that OpenSolaris
is going to (a) revolutionise the Unix world, (b) Drag Linux about
a hundred years forward, and (c) significantly improve the quality of
open source code. It _might_ actually increase the likelihood of getting
applications on Solaris x86, but that's a long shot.

> Sun needs to build faster, less costly Sparc hardware.
> They need Sparc hardware that is competitive on the desktop
> and in the server world.  They need to keep after Adobe,
> Real Networks, and other proprietary vendors and make sure
> Sparc machines don't become irrelevant.

Right now, Sparc machines on the desktop are _already_ irrelevant. 
People don't use them for word processing, people don't use them for
surfing the web, and people don't use them for playing games. Their life
on the desktop is limited to computationally intensive dedicated apps
(modelling and seismic stuff, for instance), and those people almost
invariably have a PC right beside the workstation, for everything except
their compute application. Now in that realm of hardware, you're perfectly
right--Sun needs to make faster cheaper Sparc processors, and they need to
do it soon. x86/x64 is a lost faster per dollar than Sparc, and applications
are being certified for Linux/x86 or Solaris/Sparc, with the latter waning.

Note that I'm writing this on an Ultra2, which I'm ssh'd into from my
desktop Ultra30. There are going to be a large number of people who will
use Solaris on the desktop, and will use Sparc hardware if they can. (Rich,
are you listening? :-) From a commercially viable point of view though,
Sparc desktops are dead unless Sun can completely reinvent the workstation.
(Interestingly, Bryan Cantrill hinted recently that this might be the case,
partly due to the return of Andy Bechtolsheim.)

In many ways, Sun is the last Unix company around. IBM and HP would both
barely blink if their Unices went away, DEC has been dead for ages, and
SGI has been part of the walking dead for half a decade or more. It's
getting hard to see just how a Unix company can survive, despite superior
technology. I hope they manage it.

Colin

In article <428cf819@news.nucleus.com>,
Colin B. <cbigam@somewhereelse.nucleus.com> wrote:
>(Interestingly, Bryan Cantrill hinted recently that this might be the case,
>partly due to the return of Andy Bechtolsheim.)

Has anyone seen IDC, Gartner or other's numbers for AMD64 workstations?
<URL:http://www-132.ibm.com/webapp/wcs/stores/servlet/CategoryDisplay?catalogId=-840&storeId=1&langId=-1&dualCurrId=73&categoryId=2588412>
<URL:http://www.hp.com/workstations/pws/xw9300/>
<URL:http://www.sun.com/desktop/workstation/w2100z/>

What does Sun need to make Andy Bechtolsheim upcoming workstation systems
more compelling than HP's and IBM's?

John
groenveld@acm.org

groenvel@cse.psu.edu (John D Groenveld) writes:

> In article <ic8y2bfj0u.fsf@mk.telcordia.com>,
> Dan Espen  <daneNO@SPAM.mk.telcordia.com> wrote:
>>1. The famous troll Xxxx Xxx is pretending to promote
>
> If it's a kook, then you're feeding its psychosis by naming it in
> plublic.
>
> Please don't.

Is repeating the name as bad as mentioning the name?

I admit, I have NO CLUE as to what motivates a troll.
Interesting theory.

groenvel@cse.psu.edu (John D Groenveld) writes:

> In article <ic8y2bfj0u.fsf@mk.telcordia.com>,
> Dan Espen  <daneNO@SPAM.mk.telcordia.com> wrote:
>>3. As a recent thread made so clear, thinking that opensourcing
>>Solaris is going to help Sun, is a crackheads delusion.
>
> Can you summarize or point to a good summary of that thread?

http://groups-beta.google.com/group/comp.unix.solaris/browse_thread/thread/16fd64518c609045/54696b88e008aaf8?q=sun+management+smokes+crack&rnum=1#54696b88e008aaf8

On Thu, 19 May 2005, Colin B. wrote:

> OK picture aside, I can't imagine why Sun can expect to survive a close
> development partnership with Microsoft when no other company has managed
> to do so in 30 years. You have to know that Microsoft wouldn't even bother
> dealing closely with Sun unless they can either buy or crush them.

True, but I think that M$ playing nicely was a big part of those other companies
plans.

> Right now, Sparc machines on the desktop are _already_ irrelevant.
> People don't use them for word processing, people don't use them for
> surfing the web, and people don't use them for playing games. Their life

I do!

> on the desktop is limited to computationally intensive dedicated apps
> (modelling and seismic stuff, for instance), and those people almost
> invariably have a PC right beside the workstation, for everything except

That bit I don't get.  Unless the company has deicded to look themselves
into non-standards based email system, in which case, morre fool tehm!

> desktop Ultra30. There are going to be a large number of people who will
> use Solaris on the desktop, and will use Sparc hardware if they can. (Rich,
> are you listening? :-) From a commercially viable point of view though,

Always.  :-)

> Sparc desktops are dead unless Sun can completely reinvent the workstation.

I think that before that happens, we need to see full support on Solaris x86
of all the SPARC desktop apps.  Admittedly, some of this is out of Sun's hands
(e.g., Acrobat reader, Real, etc.).

> getting hard to see just how a Unix company can survive, despite superior
> technology. I hope they manage it.

Me too.  I think if anyone can do it, it's Sun, because of their UNIX focus.

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

President,
Rite Online Inc.

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

In article <iczmuqexfb.fsf@mk.telcordia.com>,
Dan Espen  <daneNO@SPAM.mk.telcordia.com> wrote:
>http://groups-beta.google.com/group/comp.unix.solaris/browse_thread/thread/16fd64518c609045/54696b88e008aaf8?q=sun+management+smokes+crack&rnum=1#54696b88e008aaf8

This message?
<URL:http://groups.google.fi/groups?hl=en&selm=m3wts0gyg0.fsf%40athena.pienet>

I don't see any reference to Open Solaris.

John
groenveld@acm.org

> This one scares me too--especially that picture of Mcnealy and Ballmer
> on Sun's homepage. Steve's expression looks very much like a ghoul who
> has just eaten a particularly juicy human heart.

One could wonder if Sun hasn't learned a valuable lesson when MS
licensed Java, in the beginning they liked it but that mood soon turned
180 degrees. Another explenation, somehow, could be that Sun considers
that they survived the 'Java debacle' so they should be safe here too.

> I don't know about that. What I _do_ expect though, is that
> OpenSolaris is going to
> (a) revolutionise the Unix world,

Would OpenSolaris be considered Unix or a Unix-like OS ?  I don't think
it will be that revolutionair, but that heavily depends on the way it
will be implemented and managed I think.

> (b) Drag Linux about a hundred years forward,

Very doubtfull.  From what I've read about it and from what I've seen
from Solaris 10 so far I'd say that the product has a very strong
potential to make the life of Linux very *very* tough.

Its getting more and more difficult to maintain a Linux server and
seriously tell people that you won't suffer from anything nasty and that
your setup is secured. Sure, you can take a lot preventions against
local root exploits but it almost looks as if they are spreading like
wildfire.. With active development in the latest kernel I'd normally
avoid it but in these times you have little choice, unfortunatly there
pop up more bugs in there as well, some which are not even reconized as
such although some veteran kernel coders think otherwise and are now
openly critizing.

All of this sillyness was a major weighing factor for me to prepare a
shift from Debian to Solaris 10 on my server, and there are more to
follow I'm sure.

Now, with OpenSolaris and with that a more "open" community (don't take
this too literal please) might convince even more people.

> and (c) significantly improve the quality of open source code.

There is something I'm wondering about when it comes to OpenSolaris; who
will maintain the several software packages and actually make the call
on what will and what won't be involved ?

Could we expect the same quality we see in the current Solaris versions
or are people going to compromise ?


The reason I'm asking..  I've used Linux for many years now and although
I still think the OS has a very good potential I'm also wondering if its
not getting too big for its own good.. Or, in another context, if its
not being taken too serious as it was intended..

For example... On Linux you see many people who are contributing to a
certain distribution as 'package maintainer'. Which basicly means that
they package the software in a way where it will be (relatively) easy to
install. Next to installing it will (usually) also take care of some
basic settings so that it can be used/configured with a minimum effort.

So far, so good...

But you also see a tendency where people maintain software packages but
apperantly have no idea how the software really works. This can be
easily spotted when you look at packages which contain support software
for Apache (wiki's, cms', etc.).  Almost all of them will happily edit
your config file to include their own file, and issue a restart of the
server (apachectl restart). Needless to say what this will do to your
active connections, very annoying too when you consider that a mere
'apachectl graceful' would have done the trick just as well, *esp* if
you consider the time it takes to go through an entire install script.


Now I am really wondering in what way this might apply to OpenSolaris.
At this moment Solaris has a very high quality (at least in my opinion)
and I'm somewhat worried if that quality level won't decrease when more
people start applying to help out and maintain certain aspects of the
software.

Naturally this heavily depends on the whole model after which
OpenSolaris will be based and quite frankly I have to admit not having
followed that very closely.

> In many ways, Sun is the last Unix company around. IBM and HP would
> both barely blink if their Unices went away, DEC has been dead for
> ages, and SGI has been part of the walking dead for half a decade or
> more.

I agree.  Although this may sound like a very ignorant or naieve
statement I also strongly believe that Sun also cares for its customers.
For example the Solaris x86 debacle..  They intended to drop it but
didn't, from what I read the protests were an influence in that
decision.

> It's getting hard to see just how a Unix company can survive, despite
> superior technology. I hope they manage it.

*nod*.

I couldn't agree more.  And its also because of those thoughts that I'm
going to apply for a service plan once I have my (x86) Solaris server
setup and going. Partly due to the patches being available (and the ease
of upgrading) but also to support a company which (IMO) cares for
quality and its customers.

-- 
Groetjes, Peter

..\\ PGP/GPG key: http://www.catslair.org/pubkey.asc

I think there's a plot already reserved next to DEC, DG, Prime,
Apollo, et alia--although I have to say that in many ways Ken Olson
has the last laugh if you look at the development teams behind the
Windows NT OSes (Dave Cutler) and current AMD processors (Dirk
Meyers).

groenvel@cse.psu.edu (John D Groenveld) writes:

> In article <iczmuqexfb.fsf@mk.telcordia.com>,
> Dan Espen  <daneNO@SPAM.mk.telcordia.com> wrote:
>>http://groups-beta.google.com/group/comp.unix.solaris/browse_thread/thread/16fd64518c609045/54696b88e008aaf8?q=sun+management+smokes+crack&rnum=1#54696b88e008aaf8
>
> This message?
> <URL:http://groups.google.fi/groups?hl=en&selm=m3wts0gyg0.fsf%40athena.pienet>
>
> I don't see any reference to Open Solaris.

You're right, the subject of the thread is the patch management policy for
Solaris 10.

However, it's the patch management policy for Open Solaris.

"Colin B." <cbigam@somewhereelse.nucleus.com> wrote in message
news:428cf819@news.nucleus.com...
>
> Right now, Sparc machines on the desktop are _already_ irrelevant.
> People don't use them for word processing, people don't use them for
> surfing the web, and people don't use them for playing games. Their life
> on the desktop is limited to computationally intensive dedicated apps
> (modelling and seismic stuff, for instance), and those people almost
> invariably have a PC right beside the workstation, for everything except
> their compute application. Now in that realm of hardware, you're perfectly
> right--Sun needs to make faster cheaper Sparc processors, and they need to
> do it soon. x86/x64 is a lost faster per dollar than Sparc, and
applications
> are being certified for Linux/x86 or Solaris/Sparc, with the latter
waning.

Sun's partnership with Fujitsu seems to point in the direction of Sun
concentrating on the low-end CPUs, while Fujitsu (perhaps with input from
Sun) makes the server-centric processors. If Sun can pull this off, make
CPUs with less I/O capabilities but perhaps higher clock-rates, while
keeping them cheap or even cheaper, you might just have the Sparc
workstations back.

BTW, Colin, excelent post.

On 2005-05-19 20:34, Colin B. wrote:
>> Sun needs to build faster, less costly Sparc hardware. They need 
>> Sparc hardware that is competitive on the desktop and in the server 
>> world.  They need to keep after Adobe, Real Networks, and other 
>> proprietary vendors and make sure Sparc machines don't become 
>> irrelevant.
>[...]
> In many ways, Sun is the last Unix company around. IBM and HP would 
> both barely blink if their Unices went away, DEC has been dead for 
> ages, and SGI has been part of the walking dead for half a decade or 
> more. It's getting hard to see just how a Unix company can survive, 
> despite superior technology. I hope they manage it.

Sun is indeed the last Unix company around!

The quality of what they make is hopefully (as you also wished) going to 
be a good enough reason to not only survive, but continue delivering 
nice Unix-based stuff :)

On 2005-05-19 21:10, Rich Teer wrote:
> On Thu, 19 May 2005, Colin B. wrote:
>> Sparc desktops are dead unless Sun can completely reinvent the workstation.
>
> I think that before that happens, we need to see full support on Solaris x86
> of all the SPARC desktop apps.  Admittedly, some of this is out of Sun's hands
> (e.g., Acrobat reader, Real, etc.).

The source-level compatibility of software written for Solaris on SPARC 
vs. software that needs to compile on Solaris x86 is absolutely *amazing*.

Risking a guess (since I haven't seen their sources) all Adobe, Real, or 
anyone else for that matter, would have to do is simply rebuild their 
applications.  With very minimal changes, if any.

I don't think that's too hard :-)

On Sun, 22 May 2005, Giorgos Keramidas wrote:

> The source-level compatibility of software written for Solaris on SPARC
> vs. software that needs to compile on Solaris x86 is absolutely *amazing*.
>
> Risking a guess (since I haven't seen their sources) all Adobe, Real, or
> anyone else for that matter, would have to do is simply rebuild their
> applications.  With very minimal changes, if any.
>
> I don't think that's too hard :-)

Agreed, and that's why I think the fact that they haven't ported (recompiled)
them yet is unforgivable.

-- 
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" <rich.teer@rite-group.com> wrote in message
news:Pine.SOL.4.58.0505221106220.14405@zaphod...
> On Sun, 22 May 2005, Giorgos Keramidas wrote:
> > Risking a guess (since I haven't seen their sources) all Adobe, Real, or
> > anyone else for that matter, would have to do is simply rebuild their
> > applications.  With very minimal changes, if any.
> >
> > I don't think that's too hard :-)
>
> Agreed, and that's why I think the fact that they haven't ported
(recompiled)
> them yet is unforgivable.

Compiling is rarely the issue, it's the testing and regression that sucks up
the bandwidth and resources. We've had lots of different things we wanted to
try but didn't have the testing resource to implement them.

Regards,

Will Hartung
(willh@msoft.com)

"Lion-O" <nosp@m.catslair.org> wrote in message
news:slrnd8q33j.5q7.nosp@linux.intranet.lan...
> There is something I'm wondering about when it comes to OpenSolaris; who
> will maintain the several software packages and actually make the call
> on what will and what won't be involved ?

I'm hoping that OpenSolaris turns into a more BSD like effort, a "whole OS",
rather than a "Kernel" and "User Space" Linux model.

One can argue that there is little need for this kind of "open" model, since
Sun will have regular Solaris to fill that one-stop-shopping niche.

But, what I seriously don't want, is for when someone says "I run
OpenSolaris", others come back with "Which distro/kernel/patch?" questions.
I want to there to be a single OpenSolaris. If someone wants to fork it like
DragonFly BSD, then fine. To me, "Linux" is a meaningless, confusing and
very gray term and part of what makes it difficult to support. I'd like for
OpenSolaris to not go down that same path.

Regards,

Will Hartung
(willh@msoft.com)

mario wrote:

> Sun's partnership with Fujitsu seems to point in the direction of Sun
> concentrating on the low-end CPUs, while Fujitsu (perhaps with input from
> Sun) makes the server-centric processors. 

Not at all!  Search on mentions of 'Rock' and 'Sun' and you'll see that
Sun is aiming for the very high end with that processor line (both in
throughput and single-thread performance).  It's fair to say that Rock
based systems will be radically different to the systems of today,
while the Fujitsu partnership systems will be bigger/better/faster etc
versions of what we're familiar with today.

> If Sun can pull this off, make
> CPUs with less I/O capabilities but perhaps higher clock-rates, while
> keeping them cheap or even cheaper, you might just have the Sparc
> workstations back.

In the "lower" new chips end Sun is pushing Niagara - 8 cores on a
single chip with 4 hardware threads per core (so effectively 32
threads of execution on a single chip).  They're for carefully targeted
markets (rather thanbeing general purpose workhorses) particularly that
of "throughput computing".  And they do it with trivial worst-case power
consumption - high computing density with low power cost.

I suspect the next really hot workstation you may want to buy from
Sun will be the Opteron based ones (not just the current ones, which are
fast, but the upcoming range: 'sun galaxy opteron' in google).  Mind
you, there's still some upcoming sparc based desktops that I wouldn't
mind replacing my home system with.

Gavin

Not speaking for Sun, but in a marketing mood

On Mon, 23 May 2005, Will Hartung wrote:

> Compiling is rarely the issue, it's the testing and regression that sucks up
> the bandwidth and resources. We've had lots of different things we wanted to

Right.  But given the (more or less) parity between SOlaris SPARC
and Solaris x86, testing for one is more or less the same as testing
for the other.

-- 
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 <rich.teer@rite-group.com> wrote:
> On Mon, 23 May 2005, Will Hartung wrote:

>> Compiling is rarely the issue, it's the testing and regression that sucks up
>> the bandwidth and resources. We've had lots of different things we wanted to

> Right.  But given the (more or less) parity between SOlaris SPARC
> and Solaris x86, testing for one is more or less the same as testing
> for the other.

less, of course.

(you're ignoring hardware architecture)

-- 
Thomas E. Dickey
http://invisible-island.net
ftp://invisible-island.net

Rich Teer <rich.teer@rite-group.com> wrote:
> On Thu, 19 May 2005, Colin B. wrote:
> 
>> OK picture aside, I can't imagine why Sun can expect to survive a close
>> development partnership with Microsoft when no other company has managed
>> to do so in 30 years. You have to know that Microsoft wouldn't even bother
>> dealing closely with Sun unless they can either buy or crush them.
> 
> True, but I think that M$ playing nicely was a big part of those other companies
> plans.

Hmm. I would say that any company which enters a contract with another
company (especially MS, but really--ANY company whatsoever) under the
expectation that both sides will play nice is a company that's about to be
bought or bankrupted.

>> Right now, Sparc machines on the desktop are _already_ irrelevant.
>> People don't use them for word processing, people don't use them for
>> surfing the web, and people don't use them for playing games. Their life
> 
> I do!

Yes, I know. So do I. :-)

>> on the desktop is limited to computationally intensive dedicated apps
>> (modelling and seismic stuff, for instance), and those people almost
>> invariably have a PC right beside the workstation, for everything except
> 
> That bit I don't get.  Unless the company has deicded to look themselves
> into non-standards based email system, in which case, morre fool tehm!

Don't even get me started on this! Oops, too late...
As they say about standards, the best thing about them is that there are
so many to choose between.

With mail, you have a choice of the real standard (based on the RFCs), or
the piss-poor de facto pseudo-standard called Exchange. Pretty much any
large company I'm aware of has made the wrong decision, and chosen the
latter. Use Exchange and Outlook, you need Windows and a fairly fast PC.
Then of course, you need antivirus software, redundant domain controllers,
and the whole infrastructure required to support it. Unfortunately, most
companies seem to think that this is necessary.

(snip)

>> Sparc desktops are dead unless Sun can completely reinvent the workstation.
> 
> I think that before that happens, we need to see full support on Solaris x86
> of all the SPARC desktop apps.  Admittedly, some of this is out of Sun's hands
> (e.g., Acrobat reader, Real, etc.).

These ones should be fairly easy to manage, as has been discussed elsewhere.
I've filed a few bug reports with Adobe about not being able to read any
PDF files that use features beyond 5.0, on a supposedly supported platform.
The harder onese, however, are the apps that actually sell workstations: The
compute apps. Mathematica, Landmark, and um...many others. Some of these
apps don't even run properly across different Sparc workstations, without
extensive hardware-specific patching. Convincing these companies to port
to Solaris/x86 is going to take more coersion than I can imagine.

>> getting hard to see just how a Unix company can survive, despite superior
>> technology. I hope they manage it.
> 
> Me too.  I think if anyone can do it, it's Sun, because of their UNIX focus.

I hope so. When people talk about the sick state of Sun, I guess the
obvious response should be 'as compared to which other Unix company?'

Colin

In article <429359b1@news.nucleus.com>,
Colin B. <cbigam@somewhereelse.nucleus.com> wrote:
>extensive hardware-specific patching. Convincing these companies to port
>to Solaris/x86 is going to take more coersion than I can imagine.

For each of the platforms that Wolfram and Landmark support their
products, how many licensees do they have?

Assuming those vendors don't tier their platform pricing, then a good
guess is that the market for Solaris x86/x64 will need to equal at least
the smallest platform market they currently support.

John
groenveld@acm.org

On Tue, 24 May 2005, Colin B. wrote:

> Hmm. I would say that any company which enters a contract with another
> company (especially MS, but really--ANY company whatsoever) under the
> expectation that both sides will play nice is a company that's about to be
> bought or bankrupted.

Perhaps, but I think that some (many) companies will play fairly.
Perhaps I'm just too naive!

> With mail, you have a choice of the real standard (based on the RFCs), or
> the piss-poor de facto pseudo-standard called Exchange. Pretty much any
> large company I'm aware of has made the wrong decision, and chosen the
> latter. Use Exchange and Outlook, you need Windows and a fairly fast PC.
> Then of course, you need antivirus software, redundant domain controllers,
> and the whole infrastructure required to support it. Unfortunately, most
> companies seem to think that this is necessary.

That's because they are fools.

> I hope so. When people talk about the sick state of Sun, I guess the
> obvious response should be 'as compared to which other Unix company?'

Indeed!

-- 
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 <rich.teer@rite-group.com> writes:

> On Tue, 24 May 2005, Colin B. wrote:

>> With mail, you have a choice of the real standard (based on the RFCs), or
>> the piss-poor de facto pseudo-standard called Exchange. Pretty much any
>> large company I'm aware of has made the wrong decision, and chosen the
>> latter. Use Exchange and Outlook, you need Windows and a fairly fast PC.
>> Then of course, you need antivirus software, redundant domain controllers,
>> and the whole infrastructure required to support it. Unfortunately, most
>> companies seem to think that this is necessary.
>
> That's because they are fools.

Just to play a devil advocate here (even though I agree completely with you
guys), a customer chooses what he has been presented the best. If he has
been convinced (by a consultant or a salesperson or by his/her cat) that
e.g. Exchange is the best for company's email, this also means that whoever
presented alternatives to that software, didn't do a good job. Sun, or IBM,
or other companies selling services and software ...

We might be good in technical side of the issue and know a lot about HW and
SW, but, on unfamiliar ground, we still depends on a word of advise. If I
am to buy a e.g. furnace, I won't (have time to) read all technical specs
and/or technical manuals (if there are such beasts), I'll go for the one
that is presented as the best one to me ...

Or something like that.

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 2005-05-23, Rich Teer <rich.teer@rite-group.com> wrote:
> On Mon, 23 May 2005, Will Hartung wrote:
>
>> Compiling is rarely the issue, it's the testing and regression that sucks up
>> the bandwidth and resources. We've had lots of different things we wanted to
>
> Right.  But given the (more or less) parity between SOlaris SPARC
> and Solaris x86, testing for one is more or less the same as testing
> for the other.

I do not agree as changes of the hardware architecture are in many cases
likely to uncover problems with undefined or dangling pointers.

Andreas.

comp.unix.solaris@expires-on-2005-06-01.usenet.andreas-borchert.de (Andreas F. Borchert) writes:

> On 2005-05-23, Rich Teer <rich.teer@rite-group.com> wrote:
>> On Mon, 23 May 2005, Will Hartung wrote:
>>
>>> Compiling is rarely the issue, it's the testing and regression that sucks up
>>> the bandwidth and resources. We've had lots of different things we wanted to
>>
>> Right.  But given the (more or less) parity between SOlaris SPARC
>> and Solaris x86, testing for one is more or less the same as testing
>> for the other.
>
> I do not agree as changes of the hardware architecture are in many cases
> likely to uncover problems with undefined or dangling pointers.


But given that there usually is Linux x86 version available, it is to be
expected that most of the HW difference bugs have already been discovered...

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 <rich.teer@rite-group.com> writes in comp.unix.solaris:
|On Mon, 23 May 2005, Will Hartung wrote:
|
|> Compiling is rarely the issue, it's the testing and regression that sucks up
|> the bandwidth and resources. We've had lots of different things we wanted to
|
|Right.  But given the (more or less) parity between SOlaris SPARC
|and Solaris x86, testing for one is more or less the same as testing
|for the other.

It's amazing how many off-by-one buffer overflows get caught when you
move to a platform that stores pointers in the other order.   

And since the applications in question are GUI's, there's unfortunately
a bit less - for instance, testing RealPlayer on SPARC won't help you
find any issues that may crop up when it tries to use the Xv extension
to accelerate video playback (Xv is only available in Xorg on x86
right now).

-- 
________________________________________________________________________
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.

On Tue, 24 May 2005, Dragan Cvetkovic wrote:

> But given that there usually is Linux x86 version available, it is to be
> expected that most of the HW difference bugs have already been discovered...

That is what I'd had in mind.

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

President,
Rite Online Inc.

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

On 2005-05-23 05:06, Will Hartung wrote:
> "Rich Teer" <rich.teer@rite-group.com> wrote in message
> news:Pine.SOL.4.58.0505221106220.14405@zaphod...
>> On Sun, 22 May 2005, Giorgos Keramidas wrote:
>>> Risking a guess (since I haven't seen their sources) all Adobe, 
>>> Real, or anyone else for that matter, would have to do is simply 
>>> rebuild their applications.  With very minimal changes, if any.
>>>
>>> I don't think that's too hard :-)
>>
>> Agreed, and that's why I think the fact that they haven't ported 
>> (recompiled) them yet is unforgivable.
>
> Compiling is rarely the issue, it's the testing and regression that 
> sucks up the bandwidth and resources. We've had lots of different 
> things we wanted to try but didn't have the testing resource to 
> implement them.

That's very true.

Having said that, though, the Acrobat Reader and Real's streaming 
servers are, to the best of my knowledge, software-only products though, 
with few if any dependencies on particular hardware platform-specific 
requirements.  I'm hoping that regression and QA testing of such 
products won't be as great a timesink as, say, developing from scratch a 
port for any random operating system available out there :-)

Andreas F. Borchert writes:
>On 2005-05-23, Rich Teer <rich.teer@rite-group.com> wrote:
>> On Mon, 23 May 2005, Will Hartung wrote:

>>> Compiling is rarely the issue, it's the testing and regression that sucks
>>> up the bandwidth and resources. We've had lots of different things we
>>> wanted to
>>
>> Right.  But given the (more or less) parity between SOlaris SPARC and
>> Solaris x86, testing for one is more or less the same as testing for the
>> other.
>
> I do not agree as changes of the hardware architecture are in many cases
> likely to uncover problems with undefined or dangling pointers.

Actually, dangling pointers are not what I'm mostly wary of when porting
stuff from Solaris/SPARC to Solaris/x86 or other x86 systems.  It's
byte-order issues are, I'm afraid, far more often the cause of strange
problems.

Software that includes networking support and is written for big-endian
architectures is often written carelessly, assuming that the byte-order
is the same as network byte-order (which happens to be true on
big-endian architectures).  Then when the same program is compiled on
Linux/x86 or Solaris/x86, weird problems creep up.

But this is quickly getting off-topic :)

- Giorgos

Gavin Maltby wrote:
> In the "lower" new chips end Sun is pushing Niagara - 8 cores on a
> single chip with 4 hardware threads per core (so effectively 32
> threads of execution on a single chip).  They're for carefully targeted
> markets (rather thanbeing general purpose workhorses) particularly that
> of "throughput computing".  And they do it with trivial worst-case power
> consumption - high computing density with low power cost.

So, if I understood what you wrote correctly, `psrinfo -v` will return 
32 virtual processors?

> I suspect the next really hot workstation you may want to buy from
> Sun will be the Opteron based ones (not just the current ones, which are
> fast, but the upcoming range: 'sun galaxy opteron' in google).  Mind
> you, there's still some upcoming sparc based desktops that I wouldn't
> mind replacing my home system with.

I've been a long time fan of the SBUS/UPA/PCI (Ultra)SPARC architecture, 
all the way from the sun4c days, but I hate to say it: you can see the 
end of UltraSPARC.  UltraSPARC will have to withdraw to high end 
systems, where it'll eventually have to die.  Opteron coupled with 
Solaris is the way of the future, and the potential is tremendous.  As I 
type this, I now have reached the point where I have more Solaris 
servers running on x86 than (Ultra)SPARC.  Even if UltraSPARC CPUs catch 
up in performance, the systems based on these CPUs would have to have 
their price radically marked down to compete with cheap and powerful 
Opteron systems, which offer an excellent price/performance ratio.

And somehow, I don't see that happening.  And I'd love to be wrong on 
this one.

Rich Teer wrote:
>>With mail, you have a choice of the real standard (based on the RFCs), or
>>the piss-poor de facto pseudo-standard called Exchange. Pretty much any
>>large company I'm aware of has made the wrong decision, and chosen the
>>latter. Use Exchange and Outlook, you need Windows and a fairly fast PC.
>>Then of course, you need antivirus software, redundant domain controllers,
>>and the whole infrastructure required to support it. Unfortunately, most
>>companies seem to think that this is necessary.
> 
> 
> That's because they are fools.

Not necessarily fools.  Like it or not (I do), Informatics/CS is still 
an arcane art.  There are too many amateurs in the IT/CS industry posing 
as professionals.  Since IT/CS is still basically arcane, the management 
folks making the top-tier decisions refer to the amateurs that form the 
core of their IT operations.  So these businesses move to Windows 
solutions/platform because their "professionals" are most familiar with 
this type of platform, and achieve the fastest deployment results, which 
is considered very good in the management world.

Because the management isn't familiar with the intricacies of the 
technical side of things, they can't possibly fanthom what awaits them, 
and that's huge support costs and a very expensive TCO.

As for real fools, a good example is when you hear someone, a "UNIX 
administrator", speak that they are moving their infrastructure from 
AIX, HP-UX or Solaris to Linux, because "it is cheap".  These "UNIX 
administrators" are totally uninformed about the fact that a Sun Opteron 
  or an UltraSPARC server can be had for two grand.  That's foolishness; 
and again we come to my previous point, of amateurs posing as professionals.

It's a real problem these days.  Market is tough and only the top notch 
guys can get work; these amateurs are scared for their jobs and 
consequently the level of acting has been raised.  It's really bad.

On Sun, 29 May 2005, UNIX admin wrote:

> I've been a long time fan of the SBUS/UPA/PCI (Ultra)SPARC architecture,
> all the way from the sun4c days, but I hate to say it: you can see the
> end of UltraSPARC.  UltraSPARC will have to withdraw to high end
> systems, where it'll eventually have to die.  Opteron coupled with

AGreed with one caveat: Opeteron (x86) doesn't scale as well as
SPARC, nor is it likely to.  SPARC will last all the time a big
SMP server is required for certain tasks.

-- 
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, 29 May 2005, UNIX admin wrote:

> As for real fools, a good example is when you hear someone, a "UNIX
> administrator", speak that they are moving their infrastructure from
> AIX, HP-UX or Solaris to Linux, because "it is cheap".  These "UNIX
> administrators" are totally uninformed about the fact that a Sun Opteron
>   or an UltraSPARC server can be had for two grand.  That's foolishness;
> and again we come to my previous point, of amateurs posing as professionals.

Yeah, I know what you mean.

> It's a real problem these days.  Market is tough and only the top notch
> guys can get work; these amateurs are scared for their jobs and
> consequently the level of acting has been raised.  It's really bad.

Very well said...

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

President,
Rite Online Inc.

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

 X-Archive-No: Yes

"UNIX admin" <tripivceta@hotmail.com> wrote in message
news:429a2180$0$1164$5402220f@news.sunrise.ch...> >

> > That's because they are fools.
>
> Not necessarily fools.  Like it or not (I do), Informatics/CS is still
> an arcane art.  There are too many amateurs in the IT/CS industry posing
> as professionals.  Since IT/CS is still basically arcane, the management
> folks making the top-tier decisions refer to the amateurs that form the
> core of their IT operations.

We see this a lot with Linux admins.  They think that just because their
linux box (on PC Bucket hardware) hasn't crashed for 180 days with one luser
on it, that it is ready for the enterprise.  The Linux movement deluded a
large amount of amatures into thinking they were professional admins.

> Because the management isn't familiar with the intricacies of the
> technical side of things, they can't possibly fanthom what awaits them,
> and that's huge support costs and a very expensive TCO.
>
> As for real fools, a good example is when you hear someone, a "UNIX
> administrator", speak that they are moving their infrastructure from
> AIX, HP-UX or Solaris to Linux, because "it is cheap".  These "UNIX
> administrators" are totally uninformed about the fact that a Sun Opteron
>   or an UltraSPARC server can be had for two grand.  That's foolishness;
> and again we come to my previous point, of amateurs posing as
professionals.

Even worse is that those linux "professionals" try to claim that Strace or
truss is equivalent to Dtrace!


> It's a real problem these days.  Market is tough and only the top notch
> guys can get work; these amateurs are scared for their jobs and
> consequently the level of acting has been raised.  It's really bad.

Just look at all the FUD on COLA agianst OpenSolaris.  Those guys are
quaking in their boots.  I've seen a lot of interest on KLM about OS
(OpenSolaris) [Cool, a new marketing slogan for sun, OpenSolaris, the OS in
OS!].

I've had debates about Strace vs. Dtrace.  Would some expert really lay down
the differences so we can finally put the debate to an end and establish
that DTrace is vastly superiour to anything Linux offers?  I've heard so
many Amatureish linux admins claiming that Strace is good enough and that
DTrace didn't offer anything.  The reasons they say this is because they are
amatures, and have never used DTrace.  It would be good to have a PhD
systems engineer school these Linux  advocates/amatures about DTrace.

Thanks,

Mike Cox

Some malicious bastard tried to deprive future generations of your wisdom
my sticking a no-archive line in there.  Let's just strip that off, shall
we?

------------------- begin Mike Cox post ------------------
"UNIX admin" <tripivceta@hotmail.com> wrote in message
news:429a2180$0$1164$5402220f@news.sunrise.ch...> >
 
> > That's because they are fools.
>
> Not necessarily fools.  Like it or not (I do), Informatics/CS is still
> an arcane art.  There are too many amateurs in the IT/CS industry posing
> as professionals.  Since IT/CS is still basically arcane, the management
> folks making the top-tier decisions refer to the amateurs that form the
> core of their IT operations.

We see this a lot with Linux admins.  They think that just because their
linux box (on PC Bucket hardware) hasn't crashed for 180 days with one luser
on it, that it is ready for the enterprise.  The Linux movement deluded a
large amount of amatures into thinking they were professional admins.

> Because the management isn't familiar with the intricacies of the
> technical side of things, they can't possibly fanthom what awaits them,
> and that's huge support costs and a very expensive TCO.
>
> As for real fools, a good example is when you hear someone, a "UNIX
> administrator", speak that they are moving their infrastructure from
> AIX, HP-UX or Solaris to Linux, because "it is cheap".  These "UNIX
> administrators" are totally uninformed about the fact that a Sun Opteron
>   or an UltraSPARC server can be had for two grand.  That's foolishness;
> and again we come to my previous point, of amateurs posing as professionals.

Even worse is that those linux "professionals" try to claim that Strace or
truss is equivalent to Dtrace!


> It's a real problem these days.  Market is tough and only the top notch
> guys can get work; these amateurs are scared for their jobs and
> consequently the level of acting has been raised.  It's really bad.

Just look at all the FUD on COLA agianst OpenSolaris.  Those guys are
quaking in their boots.  I've seen a lot of interest on KLM about OS
(OpenSolaris) [Cool, a new marketing slogan for sun, OpenSolaris, the OS in
OS!].

I've had debates about Strace vs. Dtrace.  Would some expert really lay down
the differences so we can finally put the debate to an end and establish
that DTrace is vastly superiour to anything Linux offers?  I've heard so
many Amatureish linux admins claiming that Strace is good enough and that
DTrace didn't offer anything.  The reasons they say this is because they are
amatures, and have never used DTrace.  It would be good to have a PhD
systems engineer school these Linux  advocates/amatures about DTrace.

Thanks,

Mike Cox
------------------- end Mike Cox post ------------------


-- 
--Tim Smith

Tim Smith wrote:

>  The Linux movement deluded a
> large amount of amatures into thinking they were professional admins.

In the same way your mother deluded you into thinking you were a human
being.

-- 
Texeme Textcasting Technology
http://www.texeme.com

In article <6OudneMZc7u49QbfRVn-jg@speakeasy.net>, jabailo@texeme.com wrote:
> Tim Smith wrote:
> 
>>  The Linux movement deluded a
>> large amount of amatures into thinking they were professional admins.
> 
> In the same way your mother deluded you into thinking you were a human
> being.

Learn to read, Bailo.  (Hint: Mike Cox wrote that).


-- 
--Tim Smith

Tim Smith wrote:
> In article <6OudneMZc7u49QbfRVn-jg@speakeasy.net>, jabailo@texeme.com wrote:
> > Tim Smith wrote:
> >
> >>  The Linux movement deluded a
> >> large amount of amatures into thinking they were professional admins.
> >
> > In the same way your mother deluded you into thinking you were a human
> > being.
>
> Learn to read, Bailo.  (Hint: Mike Cox wrote that).

heh, he has been too addicted in trolling :)

Tim Smith wrote:

> In article <6OudneMZc7u49QbfRVn-jg@speakeasy.net>, jabailo@texeme.com
> wrote:
>> Tim Smith wrote:
>> 
>>>  The Linux movement deluded a
>>> large amount of amatures into thinking they were professional admins.
>> 
>> In the same way your mother deluded you into thinking you were a human
>> being.
> 
> Learn to read, Bailo.  (Hint: Mike Cox wrote that).
> 
> 


Right, just because you decide to adopt A COMPLETELY DIFFERENT AND UNHEARD
OF format for Usenet, namely:

"------------------- begin Mike Cox post ------------------"

I, and the rest of the Internet universe, are supposed to IMMEDIATELY
recognize this format.   Where's the RPC?   Where's the IEEE Standards
meeting?

OOOhhhh, nnoooo....not for TIM SMITH.  He just goes and creates a brand new,
hard to follow, quoting style and then calls anybody a TROLL who doesn't
grasp the *brilliance* of "Tim Smith" or whatever it's name is!


-- 
Texeme Textcasting Technology
http://www.texeme.com

*  *  *   Y o u r  .  S h e p h e r d  .  A q u i l a  .  D e u s  .  ( d 2
0 0 5 x x ,    d 2 0 0 4 x x ,    d 2 0 0 3 x x ,    d 2 0 0 2 x x )   *  * 
* wrote
> 
> heh, he has been too addicted in trolling :)

Get Sodded.

-- 
Texeme Textcasting Technology
http://www.texeme.com

On Mon, 30 May 2005 11:27:54 -0700, Mike Cox wrote:

> 
> 
>  X-Archive-No: Yes
> 
> "UNIX admin" <tripivceta@hotmail.com> wrote in message
> news:429a2180$0$1164$5402220f@news.sunrise.ch...> >
> 
>> > That's because they are fools.
>>
>> Not necessarily fools.  Like it or not (I do), Informatics/CS is still
>> an arcane art.  There are too many amateurs in the IT/CS industry posing
>> as professionals.  Since IT/CS is still basically arcane, the management
>> folks making the top-tier decisions refer to the amateurs that form the
>> core of their IT operations.
> 
> We see this a lot with Linux admins.  They think that just because their
> linux box (on PC Bucket hardware) hasn't crashed for 180 days with one luser
> on it, that it is ready for the enterprise.  The Linux movement deluded a
> large amount of amatures into thinking they were professional admins.

This from a guy who whines about Linux being unable to fix his
proprietary filesystem problems on his proprietary operating system and
then thinks said operating system is the greatest thing since sliced bread
because some THIRD PARTY UTILITY fixed it.


> 
> Thanks,

Anytime.

alt wrote:
> This from a guy who whines about Linux being unable to fix his
> proprietary filesystem problems on his proprietary operating system and
> then thinks said operating system is the greatest thing since sliced bread
> because some THIRD PARTY UTILITY fixed it.

Which proprietary operating system would that be?

Unless you are referring to RedHat Advanced Server, Linux isn't proprietary.

UNIX admin <tripivceta@hotmail.com> writes:

> alt wrote:
>> This from a guy who whines about Linux being unable to fix his
>> proprietary filesystem problems on his proprietary operating system and
>> then thinks said operating system is the greatest thing since sliced bread
>> because some THIRD PARTY UTILITY fixed it.
>
> Which proprietary operating system would that be?
>
> Unless you are referring to RedHat Advanced Server, Linux isn't proprietary.

I thought he meant MS Windows ...

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 Fri, 03 Jun 2005 18:35:53 +0200, UNIX admin wrote:

> 
> 
> alt wrote:
>> This from a guy who whines about Linux being unable to fix his
>> proprietary filesystem problems on his proprietary operating system and
>> then thinks said operating system is the greatest thing since sliced bread
>> because some THIRD PARTY UTILITY fixed it.
> 
> Which proprietary operating system would that be?
> 
> Unless you are referring to RedHat Advanced Server, Linux isn't proprietary.

Mike Cox started a thread in COLA where he whined about how Linux couldn't
fix his NTFS partition....

alt wrote:

> Mike Cox started a thread in COLA where he whined about how Linux couldn't
> fix his NTFS partition....

Yeah well, I don't really know if there is *anything* or *anyone* that 
can fix NTFS.

On Fri, 03 Jun 2005 18:35:53 +0200, UNIX admin wrote:

> Which proprietary operating system would that be?
> Unless you are referring to RedHat Advanced Server, Linux isn't proprietary.

explain to me why RH is proprietary. if your reason is because Sun said
so, well...

dirk
-- 
Sometimes I wonder if I'm in my right mind.  Then it passes off and I'm
as intelligent as ever.
		-- Samuel Beckett, "Endgame"

On Sun, 05 Jun 2005 20:35:12 +0200, UNIX admin wrote:

> 
> 
> alt wrote:
> 
>> Mike Cox started a thread in COLA where he whined about how Linux couldn't
>> fix his NTFS partition....
> 
> Yeah well, I don't really know if there is *anything* or *anyone* that 
> can fix NTFS.

I'm sure there's someone (and only one) in MS HQ that actually understands
NTFS and can fix it (I have a hard time believing even Microsoft is that
inept...), but other than that, yeah.... 

"dirk dierickx" <dirk.dierickx@liamg.moc> wrote in message 
news:pan.2005.06.05.22.51.09.801840@liamg.moc...
> On Fri, 03 Jun 2005 18:35:53 +0200, UNIX admin wrote:
>
>> Which proprietary operating system would that be?
>> Unless you are referring to RedHat Advanced Server, Linux isn't 
>> proprietary.
>
> explain to me why RH is proprietary. if your reason is because Sun said
> so, well...
>
> dirk
> -- 
> Sometimes I wonder if I'm in my right mind.  Then it passes off and I'm
> as intelligent as ever.
> -- Samuel Beckett, "Endgame"
>

Redhat doesnt use the standard kernel released by Linus and others on 
kernel.org, they add tons of so called (features) hense a fork, so it is 
propietary to that extent. This is same reason why most big ISV's only 
support Redhat and not other distros, just because the code is open dont 
mean you cant add propietary extnetions to it.

--
Rodrick R. Brown 

In article <DrNoe.7650$jU5.1775091@twister.nyc.rr.com>,
	Rodrick Brown <rodrick.brown@gmail.com> wrote:
> "dirk dierickx" <dirk.dierickx@liamg.moc> wrote in message 
> news:pan.2005.06.05.22.51.09.801840@liamg.moc...
>> explain to me why RH is proprietary. if your reason is because Sun said
>> so, well...
>
> Redhat doesnt use the standard kernel released by Linus and others on 
> kernel.org, they add tons of so called (features) hense a fork, so it is 
> propietary to that extent. This is same reason why most big ISV's only 
> support Redhat and not other distros, just because the code is open dont 
> mean you cant add propietary extnetions to it.

Except that the additions are also open (in the case of RedHat),
so it's a bit harsh to describe them as proprietary extensiosn.

-- 
Mike Bristow - really a very good driver

On Mon, 6 Jun 2005, Mike Bristow wrote:

> Except that the additions are also open (in the case of RedHat),
> so it's a bit harsh to describe them as proprietary extensiosn.

Like too many people, you seem to be equating "proprietory" with
"not open source".  Check out the defintions on dictionary.com;
by definition, Linux is a proprietory OS, because it is protected
by a trademark.

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

President,
Rite Online Inc.

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

In article <Pine.SOL.4.58.0506060734590.5968@zen.rite-group.com>,
	Rich Teer <rich.teer@rite-group.com> wrote:
> On Mon, 6 Jun 2005, Mike Bristow wrote:
>
>> Except that the additions are also open (in the case of RedHat),
>> so it's a bit harsh to describe them as proprietary extensiosn.
>
> Like too many people, you seem to be equating "proprietory" with
> "not open source".  Check out the defintions on dictionary.com;

I tend to think that the New Hackers Dictionary (aka, the Jargon
File) might be a better source of the terminology used by computer
folk - I doubt that the dictionary.com definition of "peek" matches
what most computer folk mean by it. [1]

Of course, the Jargon File is edited by a OSS nut, so there may be
a teeny bit of bias in that particular definition.... but my point
about "dictionaries don't (necessarily) capture the slang of a group"
still stands, I'm afraid.

But I take your point.


[1] I've just looked:  it doesn't.

-- 
Mike Bristow - really a very good driver

On Mon, 6 Jun 2005, Mike Bristow wrote:

> Of course, the Jargon File is edited by a OSS nut, so there may be
> a teeny bit of bias in that particular definition.... but my point
> about "dictionaries don't (necessarily) capture the slang of a group"
> still stands, I'm afraid.

I agree wholeheartedly with you here; I'm careful to use the canonical
definition of the word hacker, using "cracker" when I mean that.

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

President,
Rite Online Inc.

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

In comp.os.linux.advocacy, alt
<spamtrap@lazyeyez.net>
 wrote
on Mon, 06 Jun 2005 00:48:03 GMT
<pan.2005.06.06.00.45.46.532403@lazyeyez.net>:
> On Sun, 05 Jun 2005 20:35:12 +0200, UNIX admin wrote:
>
>> 
>> 
>> alt wrote:
>> 
>>> Mike Cox started a thread in COLA where he whined about how Linux couldn't
>>> fix his NTFS partition....
>> 
>> Yeah well, I don't really know if there is *anything* or *anyone* that 
>> can fix NTFS.
>
> I'm sure there's someone (and only one) in MS HQ that actually understands
> NTFS and can fix it (I have a hard time believing even Microsoft is that
> inept...), but other than that, yeah.... 
>

AIUI, the Linux kernel already has a number of the data structures
doped out so that it can read it, and had an experimental write at
one point (it's since been revamped so that the writes are much safer,
but no block allocations are allowed).

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

On Mon, 06 Jun 2005 01:14:11 +0000, Rodrick Brown wrote:

> Redhat doesnt use the standard kernel released by Linus and others on 
> kernel.org, they add tons of so called (features) hense a fork, so it is 
> propietary to that extent. This is same reason why most big ISV's only 

so does suse, mandrake/driva, ubuntu, xandros, hell they all do... (except
maybe slaskware, i'm not sure about that one)

oh, by the way, linus and andrew even stated recently it is the
distributors job to add stuff to the kernel as they see fit;

"Andrew's vision, as expressed at the summit, is that the mainline kernel
will be the fastest and most feature-rich kernel around, but not,
necessarily, the most stable. Final stabilization is to be done by
distributors (as happens now, really), but the distributors are expected
to merge their patches quickly."

http://kerneltrap.org/node/3513

all because the release model has changed dramatically since 2.6.

the changes redhat makes are, as mentioned above, features which are not
in the kernel _yet_ but can be expected soon. these are available, openly,
to anybody to tinker with and _add_ to their own kernel. hence, centos is
a complete remake of the RH server distro _including_ all the kernel
additions from redhat.

> support Redhat and not other distros, just because the code is open dont 
> mean you cant add propietary extnetions to it.

that has nothing to do with the kernel, in total. but is also targeted at
the other bundled packages like glibc and what have you. the ISV's target
specific distros for support reasons, and because no two distros ever ship
with the same versions of all the packages.

mostly they support RH as well as Suse, so you still have choice. and if
you are not using anything proprietary (like oracle or veritas stuff)
there is no reason to stay with RH.

dirk
-- 
"... an experienced, industrious, ambitious, and often quite often
picturesque liar."
		-- Mark Twain

On Mon, 06 Jun 2005 19:00:12 +0000, The Ghost In The Machine wrote:

> 
> 
> In comp.os.linux.advocacy, alt
> <spamtrap@lazyeyez.net>
>  wrote
> on Mon, 06 Jun 2005 00:48:03 GMT
> <pan.2005.06.06.00.45.46.532403@lazyeyez.net>:
>> On Sun, 05 Jun 2005 20:35:12 +0200, UNIX admin wrote:
>>
>>> 
>>> 
>>> alt wrote:
>>> 
>>>> Mike Cox started a thread in COLA where he whined about how Linux couldn't
>>>> fix his NTFS partition....
>>> 
>>> Yeah well, I don't really know if there is *anything* or *anyone* that 
>>> can fix NTFS.
>>
>> I'm sure there's someone (and only one) in MS HQ that actually understands
>> NTFS and can fix it (I have a hard time believing even Microsoft is that
>> inept...), but other than that, yeah.... 
>>
> 
> AIUI, the Linux kernel already has a number of the data structures
> doped out so that it can read it, and had an experimental write at
> one point (it's since been revamped so that the writes are much safer,
> but no block allocations are allowed).

Still, reverse engineering takes a _lot_ longer than having the spec sheet
in front of you from authorized sources and you can only guarantee 99%
accuracy, not 100%. Not that I'm condemning the linux NTFS effort, I
think it's a great testament to the Linux developers that they've come
this far with little to no documentation on NTFS data structures.

In comp.os.linux.advocacy, alt
<spamtrap@lazyeyez.net>
 wrote
on Tue, 07 Jun 2005 06:28:06 GMT
<pan.2005.06.07.06.25.06.703466@lazyeyez.net>:
> On Mon, 06 Jun 2005 19:00:12 +0000, The Ghost In The Machine wrote:
>
>> 
>> 
>> In comp.os.linux.advocacy, alt
>> <spamtrap@lazyeyez.net>
>>  wrote
>> on Mon, 06 Jun 2005 00:48:03 GMT
>> <pan.2005.06.06.00.45.46.532403@lazyeyez.net>:
>>> On Sun, 05 Jun 2005 20:35:12 +0200, UNIX admin wrote:
>>>
>>>> 
>>>> 
>>>> alt wrote:
>>>> 
>>>>> Mike Cox started a thread in COLA where he whined about how
>>>>> Linux couldn't fix his NTFS partition....
>>>> 
>>>> Yeah well, I don't really know if there is *anything* or
>>>> *anyone* that can fix NTFS.
>>>
>>> I'm sure there's someone (and only one) in MS HQ that
>>> actually understands NTFS and can fix it (I have a hard
>>> time believing even Microsoft is that inept...), but
>>>0other than that, yeah.... 
>>>
>> 
>> AIUI, the Linux kernel already has a number of the data structures
>> doped out so that it can read it, and had an experimental write at
>> one point (it's since been revamped so that the writes are much safer,
>> but no block allocations are allowed).
>
> Still, reverse engineering takes a _lot_ longer than having the spec sheet
> in front of you from authorized sources and you can only guarantee 99%
> accuracy, not 100%. Not that I'm condemning the linux NTFS effort, I
> think it's a great testament to the Linux developers that they've come
> this far with little to no documentation on NTFS data structures.
>

Agreed on all counts.  If one doesn't see a bit, it might quit.
Working, that is... :-)

And Microsoft reserves the right to change the meaning of those bits.
It's happened -- we at $EMPLOYER got bitten by some wayward handling
of a proxy issue while encapsulating WinInet.  (This when transitioning
from NT4 to 2k.)

At least with standards documentation one might get some warning.
(The only example coming to mind is the HTML -> XML switchover.)

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

On Fri, 03 Jun 2005 18:35:53 +0200, UNIX admin wrote:

> Unless you are referring to RedHat Advanced Server, Linux isn't proprietary.

Neither is RedHat...

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

Mike Bristow wrote:

> Except that the additions are also open (in the case of RedHat),
> so it's a bit harsh to describe them as proprietary extensiosn.

Open does not mean propreitary.
Proprietary is something that is owned by someone.

Mike Bristow wrote:

> I tend to think that the New Hackers Dictionary (aka, the Jargon
> File) might be a better source of the terminology used by computer
> folk - I doubt that the dictionary.com definition of "peek" matches
> what most computer folk mean by it. [1]

"New hackers dictionary", huh? Well, let's put that dictionary to a test.

Are there any entries for a "level packer" or "level decruncher" in 
there? What about trackloaders, Time Cruncher, Cross Linker, etcetera. 
If there are none, then everything suddenly becomes crystal clear  about 
how authoritative "new hackers dictonary" is.

Well?