Dear all,

does anybody know what is the maximal length of usernames on Solairs
6,7,8 and above?

Once, I was told that they are limited to 8 characters.  Is this
true?

Thanks for any explanation!

Kind regards,

Lukas
--
ruf at rawip dot org

On Tuesday 18 November 2003 9:20 pm in comp.sys.sun.admin Lukas Ruf wrote:

> Dear all,
> 
> does anybody know what is the maximal length of usernames on Solairs
> 6,7,8 and above?
> 
> Once, I was told that they are limited to 8 characters.  Is this
> true?

Thus is not a Solaris issue, it is a Unix standard.
Only the first 8 characters of a username, group, or password
are significant. Everything else should be ignored.

Be warned that janesmith is the same as janesmit

Using more than 8 characters is unsupported and often untested.
Breaking the rule can tickle all manner of obscure bugs.
For example, in HPUX 10.20 a long group name caused backups 
to mysteriously and silently fail.

If your lusers want a long email name use a mail alias.

-- 
My real address is crn (at) netunix (dot) com
WARNING all messages containing attachments or html will be silently
deleted. Send only plain text.

On Tue, 18 Nov 2003 22:55:09 +0000, Chris Newport
<me@see-my-sig.invalid> wrote:

>On Tuesday 18 November 2003 9:20 pm in comp.sys.sun.admin Lukas Ruf wrote:
>
>> Dear all,
>> 
>> does anybody know what is the maximal length of usernames on Solairs
>> 6,7,8 and above?
>> 
>> Once, I was told that they are limited to 8 characters.  Is this
>> true?
>
>Thus is not a Solaris issue, it is a Unix standard.
>Only the first 8 characters of a username, group, or password
>are significant. Everything else should be ignored.
>
>Be warned that janesmith is the same as janesmit
>
>Using more than 8 characters is unsupported and often untested.
>Breaking the rule can tickle all manner of obscure bugs.
>For example, in HPUX 10.20 a long group name caused backups 
>to mysteriously and silently fail.
>
>If your lusers want a long email name use a mail alias.

Why are you calling them lusers? Shouldn't the system accommodate the
user instead of the user accommodating the system?

Josh

On Tuesday 18 November 2003 11:19 pm in comp.sys.sun.admin Josh McKee wrote:

> On Tue, 18 Nov 2003 22:55:09 +0000, Chris Newport
> <me@see-my-sig.invalid> wrote:
> 
>>On Tuesday 18 November 2003 9:20 pm in comp.sys.sun.admin Lukas Ruf wrote:
>>
>>> Dear all,
>>> 
>>> does anybody know what is the maximal length of usernames on Solairs
>>> 6,7,8 and above?
>>> 
>>> Once, I was told that they are limited to 8 characters.  Is this
>>> true?
>>
>>Thus is not a Solaris issue, it is a Unix standard.
>>Only the first 8 characters of a username, group, or password
>>are significant. Everything else should be ignored.
>>
>>Be warned that janesmith is the same as janesmit
>>
>>Using more than 8 characters is unsupported and often untested.
>>Breaking the rule can tickle all manner of obscure bugs.
>>For example, in HPUX 10.20 a long group name caused backups
>>to mysteriously and silently fail.
>>
>>If your lusers want a long email name use a mail alias.
> 
> Why are you calling them lusers? Shouldn't the system accommodate the
> user instead of the user accommodating the system?

Patience, young grasshopper.
Enlightenment will come with experience.

-- 
My real address is crn (at) netunix (dot) com
WARNING all messages containing attachments or html will be silently
deleted. Send only plain text.

In article <e6alrvkhk3446jc15q43rqbblk9578kvp0@4ax.com>,
 Josh McKee <jtmckee@rm-bogus-ac.net> wrote:

> On Tue, 18 Nov 2003 22:55:09 +0000, Chris Newport
> <me@see-my-sig.invalid> wrote:
> 
> >On Tuesday 18 November 2003 9:20 pm in comp.sys.sun.admin Lukas Ruf wrote:
> >
> >> Dear all,
> >> 
> >> does anybody know what is the maximal length of usernames on Solairs
> >> 6,7,8 and above?
> >> 
> >> Once, I was told that they are limited to 8 characters.  Is this
> >> true?
> >
> >Thus is not a Solaris issue, it is a Unix standard.
> >Only the first 8 characters of a username, group, or password
> >are significant. Everything else should be ignored.
> >
> >Be warned that janesmith is the same as janesmit
> >
> >Using more than 8 characters is unsupported and often untested.
> >Breaking the rule can tickle all manner of obscure bugs.
> >For example, in HPUX 10.20 a long group name caused backups 
> >to mysteriously and silently fail.
> >
> >If your lusers want a long email name use a mail alias.
> 
> Why are you calling them lusers? Shouldn't the system accommodate the
> user instead of the user accommodating the system?

In this case, the users must accomodate the system, since it's a hard 
limit.  If you think it's cost effective to spend time modifying Solaris 
and all the parts affected by this to accomedate usernames, passwords 
and groups > 8 characters, then by all means have fun.  

I, personally, like to stay within standards.  The lowest common 
denominator for this namespace is 8 characters.  Some vendors support 
larger names, some only support 8 characters.  

YMMV.

-- 
DeeDee, don't press that button!  DeeDee!  NO!  Dee...

Lukas Ruf <ruf_block_spam@rawip.org> writes in comp.sys.sun.admin:
|does anybody know what is the maximal length of usernames on Solairs
|6,7,8 and above?
|Once, I was told that they are limited to 8 characters.  Is this
|true?

8 is the officially supported maximum length.  Longer ones mostly work,
but there are various known bugs and issues with them.

-- 
________________________________________________________________________
Alan Coopersmith                              alanc@alum.calberkeley.org
http://www.CSUA.Berkeley.EDU/~alanc/       aka: Alan.Coopersmith@Sun.COM
  Working for, but definitely not speaking for, Sun Microsystems, Inc.

me@see-my-sig.invalid writes:
>On Tuesday 18 November 2003 9:20 pm in comp.sys.sun.admin Lukas Ruf wrote:
>
>> does anybody know what is the maximal length of usernames on Solairs
>> 6,7,8 and above?
>> 
>> Once, I was told that they are limited to 8 characters.  Is this
>> true?
>
>Thus is not a Solaris issue, it is a Unix standard.
>

And then in a later post "me@see-my-sig.invalid" writes that some
Unix implementations support longer names, others don't.

Solaris can be said to support long usernames because nothing in
the base OS breaks on them.  The "bugs" which people refer to are
almost always things like information in the output of ps not
being in nice columns, as with short names.  I call these annoyances
rather than bugs.

Some other programs (that you download from the Internet rather than
come with Solaris) may break, but I haven't run into any yet.

It doesn't have to be a problem, as long as you can live with the
annoyances that long usernames produce.

(note I've only commented on usernames, not on passwords or other things)

  -Greg
-- 
Do NOT reply via e-mail.
Reply in the newsgroup.

Greg Andrews wrote:
>  me@see-my-sig.invalid writes:
> >On Tuesday 18 November 2003 9:20 pm in comp.sys.sun.admin Lukas Ruf
> >wrote:
> >
> >> does anybody know what is the maximal length of usernames on
> >> Solairs 6,7,8 and above?
> >>
>
>  Solaris can be said to support long usernames because nothing in
>  the base OS breaks on them.  The "bugs" which people refer to are
>  almost always things like information in the output of ps not being
>  in nice columns, as with short names.  I call these annoyances
>  rather than bugs.

Thanks for the clarification!

I do not care about cosmetic annoyances when it comes to system
reliability and stability.

>
>  Some other programs (that you download from the Internet rather
>  than come with Solaris) may break, but I haven't run into any yet.
>

As long as the base system -- i.e. more specifically: the kernel,
mountd, automountd, nis+ etc. -- is running stable, I am fine with it!

>  It doesn't have to be a problem, as long as you can live with the
>  annoyances that long usernames produce.
>

I can, definitely! ;-)

>  (note I've only commented on usernames, not on passwords or other
>  things)
>

what's the issue there?

Thanks for any further enlightenment!

-- 
Lukas Ruf <ruf at rawip dot org>

Josh McKee wrote:

> On Tue, 18 Nov 2003 22:55:09 +0000, Chris Newport
> <me@see-my-sig.invalid> wrote:
> 
>>
>>If your lusers want a long email name use a mail alias.
> 
> Why are you calling them lusers? Shouldn't the system accommodate the
> user instead of the user accommodating the system?
> 

<offtopic>An old BOFH-ism, which I think goes back to a hacked report
program (similar to finger, I expect) on an ITS system.  It went from
saying "x users logged in" when you pressed C-z, to saying "x losers logged
in".  Some users, apparently, didn't like being called losers, so the
compromise "lusers" was reached.  It stuck.

ISTR the help command on ITS was duly renamed "luser", too :)</offtopic>

I've just gone in and hacked the passwd stuff on a passing Solaris 8
machine.  If you have a luser with a stupid name (such as 'ihavelongname'),
then they *can* log in and *can* change their password.  I did so and the
updated passwd entry doesn't get truncated.  What you don't seem able to
do, however, is to log in as 'ihavelon'; the 8-character truncated version
of 'ihavelongname'.

*HOWEVER*, that's not to say you should do it.  Who knows how big the buffer
is in programs such as login, passwd, etc; usernames are designed to be 8
characters long and one shouldn't rock the boat.  Also, who knows what kind
of problems it could lead to should you require your Solaris systems to
interoperate with anyone else's UNIX.
-- 
Graham Lee
Wadham College
Oxford

On Tue, 18 Nov 2003 17:33:51 -0800, "Michael Vilain
<vilain@spamcop.net>" wrote:

>In article <e6alrvkhk3446jc15q43rqbblk9578kvp0@4ax.com>,
> Josh McKee <jtmckee@rm-bogus-ac.net> wrote:
>
>> On Tue, 18 Nov 2003 22:55:09 +0000, Chris Newport
>> <me@see-my-sig.invalid> wrote:
>> 
>> >On Tuesday 18 November 2003 9:20 pm in comp.sys.sun.admin Lukas Ruf wrote:
>> >
>> >> Dear all,
>> >> 
>> >> does anybody know what is the maximal length of usernames on Solairs
>> >> 6,7,8 and above?
>> >> 
>> >> Once, I was told that they are limited to 8 characters.  Is this
>> >> true?
>> >
>> >Thus is not a Solaris issue, it is a Unix standard.
>> >Only the first 8 characters of a username, group, or password
>> >are significant. Everything else should be ignored.
>> >
>> >Be warned that janesmith is the same as janesmit
>> >
>> >Using more than 8 characters is unsupported and often untested.
>> >Breaking the rule can tickle all manner of obscure bugs.
>> >For example, in HPUX 10.20 a long group name caused backups 
>> >to mysteriously and silently fail.
>> >
>> >If your lusers want a long email name use a mail alias.
>> 
>> Why are you calling them lusers? Shouldn't the system accommodate the
>> user instead of the user accommodating the system?
>
>In this case, the users must accomodate the system, since it's a hard 
>limit.  If you think it's cost effective to spend time modifying Solaris 
>and all the parts affected by this to accomedate usernames, passwords 
>and groups > 8 characters, then by all means have fun.

The problem I have is Chris' portrayal of the user as a loser because
they wish to use something more descriptive. I fail to see why that is
an unreasonable expectation and warrants the SA calling their
customers losers. Computers are supposed to work for us, not the other
way around.

>I, personally, like to stay within standards. The lowest common 
>denominator for this namespace is 8 characters.  Some vendors support 
>larger names, some only support 8 characters.  

And you wonder why Windows is the dominant platform. It may be
technically inferior to UNIX. But it appears to accommodate the user
in the areas that are most important to them.

Josh

On Wed, 19 Nov 2003 01:31:44 +0000, Chris Newport
<me@see-my-sig.invalid> wrote:

>On Tuesday 18 November 2003 11:19 pm in comp.sys.sun.admin Josh McKee wrote:
>
>> On Tue, 18 Nov 2003 22:55:09 +0000, Chris Newport
>> <me@see-my-sig.invalid> wrote:
>> 
>>>On Tuesday 18 November 2003 9:20 pm in comp.sys.sun.admin Lukas Ruf wrote:
>>>
>>>> Dear all,
>>>> 
>>>> does anybody know what is the maximal length of usernames on Solairs
>>>> 6,7,8 and above?
>>>> 
>>>> Once, I was told that they are limited to 8 characters.  Is this
>>>> true?
>>>
>>>Thus is not a Solaris issue, it is a Unix standard.
>>>Only the first 8 characters of a username, group, or password
>>>are significant. Everything else should be ignored.
>>>
>>>Be warned that janesmith is the same as janesmit
>>>
>>>Using more than 8 characters is unsupported and often untested.
>>>Breaking the rule can tickle all manner of obscure bugs.
>>>For example, in HPUX 10.20 a long group name caused backups
>>>to mysteriously and silently fail.
>>>
>>>If your lusers want a long email name use a mail alias.
>> 
>> Why are you calling them lusers? Shouldn't the system accommodate the
>> user instead of the user accommodating the system?
>
>Patience, young grasshopper.
>Enlightenment will come with experience.

I fail to see how this answered the question. Do you have an answer?

Josh

"Josh McKee" <jtmckee@rm-bogus-ac.net> wrote in message
news:mr7nrv8lng5tpgtdgu0immbmog8f07a4d1@4ax.com...
> On Tue, 18 Nov 2003 17:33:51 -0800, "Michael Vilain
> <vilain@spamcop.net>" wrote:
>
> And you wonder why Windows is the dominant platform. It may be
> technically inferior to UNIX. But it appears to accommodate the user
> in the areas that are most important to them.
>
> Josh

Indeed, if you forget the time staring at the blue screen and the daily
security patches you have to install , Windows is accommodating the user
very well ....
By the way , there is also a limitation of speed on the highway and this
without any disavantages for the users ?
Generraly , the most frequent problem with users , is that they forget their
login and password .
So what is the case about the 8 character limitations , as some users
allready can forget his login of one character :-))))

Pidbel

"pidbel" <pidbel@pidbel.pidbel.sdf> writes:

> Generraly , the most frequent problem with users , is that they forget their
> login and password .
> So what is the case about the 8 character limitations , as some users
> allready can forget his login of one character :-))))
>

Not to mention that they will soon start hating having to type all these
characters every time they login. I think that after some time they would
start wishing their username were (e.g.) jane and not janewithalongsurname.

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 Wednesday 19 November 2003 5:04 pm in comp.sys.sun.admin Josh McKee
wrote:

> On Wed, 19 Nov 2003 01:31:44 +0000, Chris Newport
> <me@see-my-sig.invalid> wrote:
> 
>>On Tuesday 18 November 2003 11:19 pm in comp.sys.sun.admin Josh McKee
>>wrote:
>>
>>> On Tue, 18 Nov 2003 22:55:09 +0000, Chris Newport
>>> <me@see-my-sig.invalid> wrote:
>>> 
>>>>On Tuesday 18 November 2003 9:20 pm in comp.sys.sun.admin Lukas Ruf
>>>>wrote:
>>>>
>>>>> Dear all,
>>>>> 
>>>>> does anybody know what is the maximal length of usernames on Solairs
>>>>> 6,7,8 and above?
>>>>> 
>>>>> Once, I was told that they are limited to 8 characters.  Is this
>>>>> true?
>>>>
>>>>Thus is not a Solaris issue, it is a Unix standard.
>>>>Only the first 8 characters of a username, group, or password
>>>>are significant. Everything else should be ignored.
>>>>
>>>>Be warned that janesmith is the same as janesmit
>>>>
>>>>Using more than 8 characters is unsupported and often untested.
>>>>Breaking the rule can tickle all manner of obscure bugs.
>>>>For example, in HPUX 10.20 a long group name caused backups
>>>>to mysteriously and silently fail.
>>>>
>>>>If your lusers want a long email name use a mail alias.
>>> 
>>> Why are you calling them lusers? Shouldn't the system accommodate the
>>> user instead of the user accommodating the system?
>>
>>Patience, young grasshopper.
>>Enlightenment will come with experience.
> 
> I fail to see how this answered the question. Do you have an answer?

The clue has been imparted - 8 characters is the maximum.
This is a hard limit, resistance is futile.
The enlightened sysadmin does not mess with the standards.
On that path lies danger.

Do you wish to absorb the clue as a step on your path to 
enlightenment or are you waiting for someone to impart it
with a cluebat after you screw something up ?.


-- 
My real address is crn (at) netunix (dot) com
WARNING all messages containing attachments or html will be silently
deleted. Send only plain text.

Josh McKee <jtmckee@rm-bogus-ac.net> writes in comp.sys.sun.admin:
|The problem I have is Chris' portrayal of the user as a loser because
|they wish to use something more descriptive. I fail to see why that is
|an unreasonable expectation and warrants the SA calling their
|customers losers. Computers are supposed to work for us, not the other
|way around.

Don't think of it as a limitation, think of it as an aid in reducing
carpal tunnel syndrome.  Why would users want to type even longer
usernames everytime they login?

-- 
________________________________________________________________________
Alan Coopersmith                              alanc@alum.calberkeley.org
http://www.CSUA.Berkeley.EDU/~alanc/       aka: Alan.Coopersmith@Sun.COM
  Working for, but definitely not speaking for, Sun Microsystems, Inc.

On Wednesday 19 November 2003 4:54 pm in comp.sys.sun.admin Josh McKee
wrote:

 
>>> Why are you calling them lusers? Shouldn't the system accommodate the
>>> user instead of the user accommodating the system?
>>
>>In this case, the users must accomodate the system, since it's a hard
>>limit.  If you think it's cost effective to spend time modifying Solaris
>>and all the parts affected by this to accomedate usernames, passwords
>>and groups > 8 characters, then by all means have fun.
> 
> The problem I have is Chris' portrayal of the user as a loser because
> they wish to use something more descriptive. I fail to see why that is
> an unreasonable expectation and warrants the SA calling their
> customers losers. Computers are supposed to work for us, not the other
> way around.

Oh purleeeese !

Are you lysdexic ?.
Can you tell the difference between luser and loser ?
The two terms are not the same, although they often overlap.

Your duty as a sysadmin is to defend the system from the 
foolishness of the lusers who use the system but lack
the clue levels required to ensure optimum reliability and
tranquility. You should be the keeper of the clue, imparting it
to those acolytes who aspire to enlightenment. You are also
the BOFH, administering the LART to those impervious to the clue.

In the cited case, a luser wants to compromise the system by
breaking the standards. It is your duty to defend the system
by not pandering to such foolishness. You are not paid to be
popular, you are paid to ensure the reliability of the system.
The simple solution is to educate the luser and impart the clue
that his email address need not be the same as his login name.
Offer him an alias, subject to RFC compliance and whatever
company standards may apply.
Think of this as a full name and a short name.

-- 
My real address is crn (at) netunix (dot) com
WARNING all messages containing attachments or html will be silently
deleted. Send only plain text.

gerg@panix.com (Greg Andrews) writes in comp.sys.sun.admin:
|Solaris can be said to support long usernames because nothing in
|the base OS breaks on them.

But Sun does not guarantee they will work, and you may not be able to
get bugs fixed if you find programs they don't work with.  If you have
a support contract and think this should change, ask to be added to the
list of customers requesting RFE 4109819.

-- 
________________________________________________________________________
Alan Coopersmith                              alanc@alum.calberkeley.org
http://www.CSUA.Berkeley.EDU/~alanc/       aka: Alan.Coopersmith@Sun.COM
  Working for, but definitely not speaking for, Sun Microsystems, Inc.

gerg@panix.com (Greg Andrews) writes:

>Solaris can be said to support long usernames because nothing in
>the base OS breaks on them.  The "bugs" which people refer to are
>almost always things like information in the output of ps not
>being in nice columns, as with short names.  I call these annoyances
>rather than bugs.

The "passwd" command breaks.  Perhaps this only happens with NIS+.

You can get around the problem with

	passwd $USER

Lukas Ruf <ruf_block_spam@rawip.org> writes:
>Greg Andrews wrote:

>>  Solaris can be said to support long usernames because nothing in
>>  the base OS breaks on them.  The "bugs" which people refer to are
>>  almost always things like information in the output of ps not being
>>  in nice columns, as with short names.  I call these annoyances
>>  rather than bugs.

>Thanks for the clarification!

>I do not care about cosmetic annoyances when it comes to system
>reliability and stability.

>As long as the base system -- i.e. more specifically: the kernel,
>mountd, automountd, nis+ etc. -- is running stable, I am fine with it!

>>  (note I've only commented on usernames, not on passwords or other
>>  things)

>what's the issue there?

I am finding that when users give a password longer than 8 characters
(when changing passwords), it breaks.  The password change itself is
fine (uses the first 8 characters).  But the NIS+ credentials seem to
break.

Neil W Rickert <rickert+nn@cs.niu.edu> writes:

>I am finding that when users give a password longer than 8 characters
>(when changing passwords), it breaks.  The password change itself is
>fine (uses the first 8 characters).  But the NIS+ credentials seem to
>break.

That's a bug; (OS release/patch version?)

Casper

On Wed, 19 Nov 2003 17:26:01 +0000, Chris Newport
<me@see-my-sig.invalid> wrote:

>On Wednesday 19 November 2003 5:04 pm in comp.sys.sun.admin Josh McKee
>wrote:
>
>> On Wed, 19 Nov 2003 01:31:44 +0000, Chris Newport
>> <me@see-my-sig.invalid> wrote:
>> 
>>>On Tuesday 18 November 2003 11:19 pm in comp.sys.sun.admin Josh McKee
>>>wrote:
>>>
>>>> On Tue, 18 Nov 2003 22:55:09 +0000, Chris Newport
>>>> <me@see-my-sig.invalid> wrote:
>>>> 
>>>>>On Tuesday 18 November 2003 9:20 pm in comp.sys.sun.admin Lukas Ruf
>>>>>wrote:
>>>>>
>>>>>> Dear all,
>>>>>> 
>>>>>> does anybody know what is the maximal length of usernames on Solairs
>>>>>> 6,7,8 and above?
>>>>>> 
>>>>>> Once, I was told that they are limited to 8 characters.  Is this
>>>>>> true?
>>>>>
>>>>>Thus is not a Solaris issue, it is a Unix standard.
>>>>>Only the first 8 characters of a username, group, or password
>>>>>are significant. Everything else should be ignored.
>>>>>
>>>>>Be warned that janesmith is the same as janesmit
>>>>>
>>>>>Using more than 8 characters is unsupported and often untested.
>>>>>Breaking the rule can tickle all manner of obscure bugs.
>>>>>For example, in HPUX 10.20 a long group name caused backups
>>>>>to mysteriously and silently fail.
>>>>>
>>>>>If your lusers want a long email name use a mail alias.
>>>> 
>>>> Why are you calling them lusers? Shouldn't the system accommodate the
>>>> user instead of the user accommodating the system?
>>>
>>>Patience, young grasshopper.
>>>Enlightenment will come with experience.
>> 
>> I fail to see how this answered the question. Do you have an answer?
>
>The clue has been imparted - 8 characters is the maximum.

I am well aware of this.

>This is a hard limit, resistance is futile.

Resistance? What resistance are you referring to?

>The enlightened sysadmin does not mess with the standards.

Who said anything about messing with standards?

>On that path lies danger.
>
>Do you wish to absorb the clue as a step on your path to 
>enlightenment or are you waiting for someone to impart it
>with a cluebat after you screw something up ?.

Who said anything about messing with the standard? I am merely
inquiring into your classification of users who want longer than 8
characters in their user ID as losers. I don't think that it is
unreasonable for a user to want to use what works best for them and
not what works best for the system. Computers should work for the
user, not the other way around. If the standard is limiting perhaps
the standard should be altered to accommodate the users requirements.

Josh

On Wed, 19 Nov 2003 18:08:12 +0100, "pidbel"
<pidbel@pidbel.pidbel.sdf> wrote:

>
>"Josh McKee" <jtmckee@rm-bogus-ac.net> wrote in message
>news:mr7nrv8lng5tpgtdgu0immbmog8f07a4d1@4ax.com...
>> On Tue, 18 Nov 2003 17:33:51 -0800, "Michael Vilain
>> <vilain@spamcop.net>" wrote:
>>
>> And you wonder why Windows is the dominant platform. It may be
>> technically inferior to UNIX. But it appears to accommodate the user
>> in the areas that are most important to them.
>>
>> Josh
>
>Indeed, if you forget the time staring at the blue screen and the daily
>security patches you have to install , Windows is accommodating the user
>very well ....

Yet Windows still commands a huge share of the market. Why is that?
It's because Windows is giving the user what they want. It should be
obvious that ease of use is what the users want. Reliability and
security are taking a back seat to user friendliness. You can present
your case about the strengths of UNIX and the weaknesses of Windows
all day long. But in the end the user is going to want the user
friendliness of Windows over the reliability of UNIX.

>By the way , there is also a limitation of speed on the highway and this
>without any disavantages for the users ?
>Generraly , the most frequent problem with users , is that they forget their
>login and password .

They probably forget their user ID because it's something cryptic
designed to fit within an 8 character limitation of the system. It
requires the user to adapt to the system and not the other way around.

>So what is the case about the 8 character limitations , as some users
>allready can forget his login of one character :-))))

But if it were their name how likely do you think that they'd forget
it? Probably unlikely.

Josh

On Wed, 19 Nov 2003 23:23:00 +0000, Chris Newport
<me@see-my-sig.invalid> wrote:

>On Wednesday 19 November 2003 4:54 pm in comp.sys.sun.admin Josh McKee
>wrote:
>
> 
>>>> Why are you calling them lusers? Shouldn't the system accommodate the
>>>> user instead of the user accommodating the system?
>>>
>>>In this case, the users must accomodate the system, since it's a hard
>>>limit.  If you think it's cost effective to spend time modifying Solaris
>>>and all the parts affected by this to accomedate usernames, passwords
>>>and groups > 8 characters, then by all means have fun.
>> 
>> The problem I have is Chris' portrayal of the user as a loser because
>> they wish to use something more descriptive. I fail to see why that is
>> an unreasonable expectation and warrants the SA calling their
>> customers losers. Computers are supposed to work for us, not the other
>> way around.
>
>Oh purleeeese !
>
>Are you lysdexic ?.
>Can you tell the difference between luser and loser ?

No.

>The two terms are not the same, although they often overlap.

Perhaps you'd be willing to explain why loser and luser are not, for
all intents and purposes, the same? According to dictionary.com they
appear to be synonymous.

>Your duty as a sysadmin is to defend the system from the 
>foolishness of the lusers who use the system but lack
>the clue levels required to ensure optimum reliability and
>tranquility.

I'm not disagreeing with your obligation as a system admin. What I am
disagreeing with is your classification of the user as a loser.
Especially when they may not be aware of the limitation. It's your job
to know...not theirs.

>You should be the keeper of the clue, imparting it
>to those acolytes who aspire to enlightenment. You are also
>the BOFH, administering the LART to those impervious to the clue.
>
>In the cited case, a luser wants to compromise the system by
>breaking the standards.

What post are you referring to? The OP was making an inquiry as to
what the standard was. I saw nothing in the post that indicated
*anyone* wanted to break the standard.

>It is your duty to defend the system by not pandering to such foolishness.

We're not in disagreement on this. What is your point?

>You are not paid to be popular, you are paid to ensure the reliability of
>the system.

But, IMO, you should not go around calling people lusers because they
probably aren't even aware that there is a limitation.

>The simple solution is to educate the luser and impart the clue
>that his email address need not be the same as his login name.

Nothing wrong with that...except that you're classifying them as
losers instead of users. I disagree with that. Especially when the
computer should accommodate the user and not the user having to
accommodate the system. Do you consider that unreasonable?

Josh

>Offer him an alias, subject to RFC compliance and whatever
>company standards may apply.
>Think of this as a full name and a short name.

In article <e5kprvg380asilftdn2dpustmml3hkpi43@4ax.com>,
Josh McKee  <jtmckee@rm-bogus-ac.net> wrote:

> Yet Windows still commands a huge share of the market. Why is that?
> It's because Windows is giving the user what they want.

Can you back that statement with some proof? A far more likely
explanation for the market share is tha fact that it is installed
at the factory on practically every PC sold.

Before MS Windows existed MS-DOS/PC-DOS commanded a huge share
of the market. It was also installed at the factory on practically
every PC sold at that time. No one can say that MS-DOS/PC-DOS gave
the user what they wanted, after all it was abandoned as soon as
MS Windows became installed at the factory.

-- 
G�ran Larsson     http://www.mitt-eget.com/

On Thu, 20 Nov 2003 14:53:41 GMT, hoh@invalid.invalid (Goran Larsson)
wrote:

>In article <e5kprvg380asilftdn2dpustmml3hkpi43@4ax.com>,
>Josh McKee  <jtmckee@rm-bogus-ac.net> wrote:
>
>> Yet Windows still commands a huge share of the market. Why is that?
>> It's because Windows is giving the user what they want.
>
>Can you back that statement with some proof?

I don't have any actual numbers but I think it should be obvious that
people are going to prefer the user friendliness of Windows over UNIX.
This should go without saying.

>A far more likely explanation for the market share is tha fact that it is installed
>at the factory on practically every PC sold.
>
>Before MS Windows existed MS-DOS/PC-DOS commanded a huge share
>of the market. It was also installed at the factory on practically
>every PC sold at that time. No one can say that MS-DOS/PC-DOS gave
>the user what they wanted, after all it was abandoned as soon as
>MS Windows became installed at the factory.

My assertion is that Windows is offering more of what users want. The
abandonment of MS-DOS to Windows would appear to support that
assertion. Windows offered more of what users wanted over MS-DOS.

Josh

Josh McKee wrote:

> On Wed, 19 Nov 2003 23:23:00 +0000, Chris Newport
> <me@see-my-sig.invalid> wrote:
> 
>>On Wednesday 19 November 2003 4:54 pm in comp.sys.sun.admin Josh McKee
>>wrote:
>>
>> 
>>>>> Why are you calling them lusers? Shouldn't the system accommodate the
>>>>> user instead of the user accommodating the system?
>>>>
>>>>In this case, the users must accomodate the system, since it's a hard
>>>>limit.  If you think it's cost effective to spend time modifying Solaris
>>>>and all the parts affected by this to accomedate usernames, passwords
>>>>and groups > 8 characters, then by all means have fun.
>>> 
>>> The problem I have is Chris' portrayal of the user as a loser because
>>> they wish to use something more descriptive. I fail to see why that is
>>> an unreasonable expectation and warrants the SA calling their
>>> customers losers. Computers are supposed to work for us, not the other
>>> way around.
>>
>>Oh purleeeese !
>>
>>Are you lysdexic ?.
>>Can you tell the difference between luser and loser ?
> 
> No.
> 
>>The two terms are not the same, although they often overlap.
> 
> Perhaps you'd be willing to explain why loser and luser are not, for
> all intents and purposes, the same? According to dictionary.com they
> appear to be synonymous.
> 

[pasted from the time I explained this earlier in the thread]
> Why are you calling them lusers? Shouldn't the system accommodate the
> user instead of the user accommodating the system?
> 

<offtopic>An old BOFH-ism, which I think goes back to a hacked report
program (similar to finger, I expect) on an ITS system.  It went from
saying "x users logged in" when you pressed C-z, to saying "x losers logged
in".  Some users, apparently, didn't like being called losers, so the
compromise "lusers" was reached.  It stuck.

ISTR the help command on ITS was duly renamed "luser", too :)</offtopic>


A luser is therefore somewhere intermediate to loser and user.  Perhaps a
standard user, who lacks the clue/patience/ability/wherewithal to find the
answers to trivial questions themselves, but is not sufficiently disliked
to become a loser.  Or one who is quite productive, therefore not a loser,
but has *some* loser-ish traits, such as relying on getpwpit(3)[*].  Luser
is often used as a term of endearment in the pub of an evening; whereas
loser is not.

[*] NAME
        getpwpit() - get password entry

SYNOPSIS
        char *getpwpit(postit *pit,monitor *mon);

DESCRIPTION
        getpwpit() retrieves the unencrypted password for the luser whose monitor
is pointed to by mon.  The password itself resides on the post-it note
pointed to by pit.

RETURN VALUE
        getpwpit() returns the UNencrypted password stored on the post-it note.  If
no post-it note can be found, getpwpit() will return NULL and set errno
depending on the error that occurred.

ERRORS
        ENOPIT No post-it exists at the monitor specified.  This usually results in
a call to sysadmind(8).
        ELART  The luser has been readjusted so that the password is no longer
stored on a post-it note attached to their monitor.

BUGS
        ELART is often not considered to be an error :)
        getpwpit() is not thought of as a secure password access method.
-- 
Graham Lee
Wadham College
Oxford

"Josh McKee" <jtmckee@rm-bogus-ac.net> wrote in message
news:e5kprvg380asilftdn2dpustmml3hkpi43@4ax.com...
> On Wed, 19 Nov 2003 18:08:12 +0100, "pidbel"
> <pidbel@pidbel.pidbel.sdf> wrote:
..> Yet Windows still commands a huge share of the market. Why is that?

It is not because you have the bigest that you are the best  .....


> It's because Windows is giving the user what they want. It should be
> obvious that ease of use is what the users want. Reliability and
> security are taking a back seat to user friendliness. You can present
> your case about the strengths of UNIX and the weaknesses of Windows
> all day long. But in the end the user is going to want the user
> friendliness of Windows over the reliability of UNIX.
>
Windows is only given users what Bill Gates want to give to them . If you
want to add something to the
system you cannot as the code is still confidential ( think about the
complaint of the CEE against Microsoft ).
By the way , windows is only an application which is Gates call an operating
system .
Here I say it again : blue screen . Ask the users around you and I think you
will agree if you are not staying the whole day at your manager's desk .

> >By the way , there is also a limitation of speed on the highway and this
> >without any disavantages for the users ?
> >Generraly , the most frequent problem with users , is that they forget
their
> >login and password .
>
> They probably forget their user ID because it's something cryptic
> designed to fit within an 8 character limitation of the system. It
> requires the user to adapt to the system and not the other way around.
>
> >So what is the case about the 8 character limitations , as some users
> >allready can forget his login of one character :-))))
>
> But if it were their name how likely do you think that they'd forget
> it? Probably unlikely.

Did you allready read a book about computer security and passwords .
OK , I understand now why you have difficulties with Unix and other real
operating systems .
I hope you are not working at a Bank , otherwise you lost now allready
thousands of dollars with your name
>
> Josh

pidbel wrote:

> 
> "Josh McKee" <jtmckee@rm-bogus-ac.net> wrote in message
> news:e5kprvg380asilftdn2dpustmml3hkpi43@4ax.com...
>> On Wed, 19 Nov 2003 18:08:12 +0100, "pidbel"
>> <pidbel@pidbel.pidbel.sdf> wrote:
> .> Yet Windows still commands a huge share of the market. Why is that?
> 
> It is not because you have the bigest that you are the best  .....
> 
> 
>> It's because Windows is giving the user what they want. It should be
>> obvious that ease of use is what the users want. Reliability and
>> security are taking a back seat to user friendliness. You can present
>> your case about the strengths of UNIX and the weaknesses of Windows
>> all day long. But in the end the user is going to want the user
>> friendliness of Windows over the reliability of UNIX.
>>
> Windows is only given users what Bill Gates want to give to them . If you
> want to add something to the
> system you cannot as the code is still confidential ( think about the
> complaint of the CEE against Microsoft ).

The Solaris code is also confidential, isn't it?  As was SunOS', IIRC.
-- 
Graham Lee
Wadham College
Oxford

On Thu, 20 Nov 2003, Graham Lee wrote:

> [*] NAME
>         getpwpit() - get password entry
>
> SYNOPSIS
>         char *getpwpit(postit *pit,monitor *mon);
>
> DESCRIPTION
>         getpwpit() retrieves the unencrypted password for the luser whose monitor
> is pointed to by mon.  The password itself resides on the post-it note
> pointed to by pit.
>
> RETURN VALUE
>         getpwpit() returns the UNencrypted password stored on the post-it note.  If
> no post-it note can be found, getpwpit() will return NULL and set errno
> depending on the error that occurred.
>
> ERRORS
>         ENOPIT No post-it exists at the monitor specified.  This usually results in
> a call to sysadmind(8).
>         ELART  The luser has been readjusted so that the password is no longer
> stored on a post-it note attached to their monitor.
>
> BUGS
>         ELART is often not considered to be an error :)
>         getpwpit() is not thought of as a secure password access method.

Now THAT is funny!  :-)

-- 
Rich Teer, SCNA, SCSA

President,
Rite Online Inc.

Voice: +1 (250) 979-1638
URL: http://www.rite-online.net

Graham Lee <graham.lee@wadham.oxford.ac.invalid.uk> writes:

>> Windows is only given users what Bill Gates want to give to them . If you
>> want to add something to the
>> system you cannot as the code is still confidential ( think about the
>> complaint of the CEE against Microsoft ).
>
> The Solaris code is also confidential, isn't it?  As was SunOS', IIRC.

Not if you are an university. And Solaris 8 source code was available for
some time. 

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 <7nkprv4g4iqm593grkhvq84rven8i9fap5@4ax.com>,
Josh McKee  <jtmckee@rm-bogus-ac.net> did thusly spew forth: 
>
>Nothing wrong with that...except that you're classifying them as
>losers instead of users. I disagree with that. Especially when the
>computer should accommodate the user and not the user having to
>accommodate the system. Do you consider that unreasonable?

What you call "height challenged" some of us call "midget". So be it.

Besides, the term goes back to a time when it was common to call
functional design problems "lossage" or to characterize something
that was in need of repair as "losing". You take a user who has 
a problem (i.e., something that needs fixed), you have a "losing
user", or l'user or "luser". "Luser" is not fundamentally the same
as a "loser". However, a "luser" transitions to a "loser" by having
way too many problems and an overall disposition of helplessness.

-tom

dragan_usenet@gmx.net writes in comp.sys.sun.admin:
|Graham Lee <graham.lee@wadham.oxford.ac.invalid.uk> writes:
|
|>> Windows is only given users what Bill Gates want to give to them . If you
|>> want to add something to the
|>> system you cannot as the code is still confidential ( think about the
|>> complaint of the CEE against Microsoft ).
|>
|> The Solaris code is also confidential, isn't it?  As was SunOS', IIRC.
|
|Not if you are an university. And Solaris 8 source code was available for
|some time. 

The license for both of those still require you to keep the bits
confidential and not share with others.

-- 
________________________________________________________________________
Alan Coopersmith                              alanc@alum.calberkeley.org
http://www.CSUA.Berkeley.EDU/~alanc/       aka: Alan.Coopersmith@Sun.COM
  Working for, but definitely not speaking for, Sun Microsystems, Inc.

Rich Teer wrote:
> On Thu, 20 Nov 2003, Graham Lee wrote:
> 
> 
>>[*] NAME
>>        getpwpit() - get password entry
>>
>>SYNOPSIS
>>        char *getpwpit(postit *pit,monitor *mon);
>>
>>DESCRIPTION
>>        getpwpit() retrieves the unencrypted password for the luser whose monitor
>>is pointed to by mon.  The password itself resides on the post-it note
>>pointed to by pit.
>>
>>RETURN VALUE
>>        getpwpit() returns the UNencrypted password stored on the post-it note.  If
>>no post-it note can be found, getpwpit() will return NULL and set errno
>>depending on the error that occurred.
>>
>>ERRORS
>>        ENOPIT No post-it exists at the monitor specified.  This usually results in
>>a call to sysadmind(8).
>>        ELART  The luser has been readjusted so that the password is no longer
>>stored on a post-it note attached to their monitor.
>>
>>BUGS
>>        ELART is often not considered to be an error :)
>>        getpwpit() is not thought of as a secure password access method.
> 
> 
> Now THAT is funny!  :-)
> 

I once wrote a Work Instruction for our operators on how they could 
locate their arses, using documented hand/arm movements.

In article <LKWdnW8V9NemYCGiRVn-ig@speakeasy.net>,
 Thomas H Jones II <ferric@xanthia.com> wrote:

> In article <7nkprv4g4iqm593grkhvq84rven8i9fap5@4ax.com>,
> Josh McKee  <jtmckee@rm-bogus-ac.net> did thusly spew forth: 
> >
> >Nothing wrong with that...except that you're classifying them as
> >losers instead of users. I disagree with that. Especially when the
> >computer should accommodate the user and not the user having to
> >accommodate the system. Do you consider that unreasonable?
> 
> What you call "height challenged" some of us call "midget". So be it.

I fail to see the relevance. Could you please clarify what you're trying 
to say?
 
> Besides, the term goes back to a time when it was common to call
> functional design problems "lossage" or to characterize something
> that was in need of repair as "losing". You take a user who has 
> a problem (i.e., something that needs fixed), you have a "losing
> user", or l'user or "luser". "Luser" is not fundamentally the same
> as a "loser". However, a "luser" transitions to a "loser" by having
> way too many problems and an overall disposition of helplessness.

Spin it any way that you want. The word "luser" appears to be synonymous 
with "loser".

Josh

In article <bpiqs3$cfb$1@news-reader1.wanadoo.fr>,
 "pidbel" <pidbel@pidbel.pidbel.sdf> wrote:

> "Josh McKee" <jtmckee@rm-bogus-ac.net> wrote in message
> news:e5kprvg380asilftdn2dpustmml3hkpi43@4ax.com...
> > On Wed, 19 Nov 2003 18:08:12 +0100, "pidbel"
> > <pidbel@pidbel.pidbel.sdf> wrote:
> .> Yet Windows still commands a huge share of the market. Why is that?
> 
> It is not because you have the bigest that you are the best  .....

That depends on how you define "best". What is best?

> > It's because Windows is giving the user what they want. It should be
> > obvious that ease of use is what the users want. Reliability and
> > security are taking a back seat to user friendliness. You can present
> > your case about the strengths of UNIX and the weaknesses of Windows
> > all day long. But in the end the user is going to want the user
> > friendliness of Windows over the reliability of UNIX.
> >
> Windows is only given users what Bill Gates want to give to them .

It appears that is giving them what they want. Thus the reason for 
Microsoft's huge market share.

> If you want to add something to the
> system you cannot as the code is still confidential ( think about the
> complaint of the CEE against Microsoft ).

I don't think that most users want to add something to the OS. They want 
someone to provide them with an OS that works the way they want. And it 
appears that Microsoft has done a reasonable job of delivering that.

> By the way , windows is only an application which is Gates call an operating
> system .
> Here I say it again : blue screen . Ask the users around you and I think you
> will agree if you are not staying the whole day at your manager's desk .

I'm not here to debate the strengths and weaknesses of Windows. Please 
do try and stay on the subject at hand. I'm merely stating that despite 
it's technical flaws Windows appears to be giving users what they want. 
And what they want appears to be a user friendly OS that is not as 
reliable over a less user friendly OS that is very reliable.

> > >By the way , there is also a limitation of speed on the highway and this
> > >without any disavantages for the users ?
> > >Generraly , the most frequent problem with users , is that they forget
> their
> > >login and password .
> >
> > They probably forget their user ID because it's something cryptic
> > designed to fit within an 8 character limitation of the system. It
> > requires the user to adapt to the system and not the other way around.
> >
> > >So what is the case about the 8 character limitations , as some users
> > >allready can forget his login of one character :-))))
> >
> > But if it were their name how likely do you think that they'd forget
> > it? Probably unlikely.
> 
> Did you allready read a book about computer security and passwords .

Non-issue. My comments aren't made with the idea of security in mind. 
That's not the focus of this discussion.

> OK , I understand now why you have difficulties with Unix and other real
> operating systems .

Yeah, I get that a lot from people in this newsgroup and then eventually 
end up explaining UNIX to them. Kind of funny if you ask me.

> I hope you are not working at a Bank , otherwise you lost now allready
> thousands of dollars with your name

Strawman. Get back to me when you can address the topic being discussed.

Josh

Josh McKee wrote:

> In article <LKWdnW8V9NemYCGiRVn-ig@speakeasy.net>,
>  Thomas H Jones II <ferric@xanthia.com> wrote:
> 
>> Besides, the term goes back to a time when it was common to call
>> functional design problems "lossage" or to characterize something
>> that was in need of repair as "losing". You take a user who has
>> a problem (i.e., something that needs fixed), you have a "losing
>> user", or l'user or "luser". "Luser" is not fundamentally the same
>> as a "loser". However, a "luser" transitions to a "loser" by having
>> way too many problems and an overall disposition of helplessness.
> 
> Spin it any way that you want. The word "luser" appears to be synonymous
> with "loser".
> 

Only if you deliberately fail to take into account all of the explanations
about their differences given in this froup as well as elsewhere.  Which
you're perfectly entitled to do of course; just don't expect everyone else
to do it too ;).
-- 
Graham Lee
Wadham College
Oxford

Rich Teer wrote:

> On Thu, 20 Nov 2003, Graham Lee wrote:
> 
>> [*] NAME
>>         getpwpit() - get password entry
>>
> 
> Now THAT is funny!  :-)
> 
Thanks :).  I was thinking; is there a newsgroup or similar where 'manpages'
such as this have a home, so that more like-minded people can read them?  I
had a quick look at the bofh.* hierarchy but nothing seemed to spring out.

Cheers.
-- 
Graham Lee
Wadham College
Oxford

On Fri, 21 Nov 2003, Graham Lee wrote:

> Thanks :).  I was thinking; is there a newsgroup or similar where 'manpages'
> such as this have a home, so that more like-minded people can read them?  I
> had a quick look at the bofh.* hierarchy but nothing seemed to spring out.

rec.humour might be one of them.  But I think that sysadmin type
humour is OK here, provided that it isn't too often.

-- 
Rich Teer, SCNA, SCSA

President,
Rite Online Inc.

Voice: +1 (250) 979-1638
URL: http://www.rite-online.net

In article <bplro6$a6p$1@news.ox.ac.uk>,
 Graham Lee <graham.lee@wadham.oxford.ac.invalid.uk> wrote:

> Josh McKee wrote:
> 
> > In article <LKWdnW8V9NemYCGiRVn-ig@speakeasy.net>,
> >  Thomas H Jones II <ferric@xanthia.com> wrote:
> > 
> >> Besides, the term goes back to a time when it was common to call
> >> functional design problems "lossage" or to characterize something
> >> that was in need of repair as "losing". You take a user who has
> >> a problem (i.e., something that needs fixed), you have a "losing
> >> user", or l'user or "luser". "Luser" is not fundamentally the same
> >> as a "loser". However, a "luser" transitions to a "loser" by having
> >> way too many problems and an overall disposition of helplessness.
> > 
> > Spin it any way that you want. The word "luser" appears to be synonymous
> > with "loser".
> > 
> 
> Only if you deliberately fail to take into account all of the explanations
> about their differences given in this froup as well as elsewhere.  Which
> you're perfectly entitled to do of course; just don't expect everyone else
> to do it too ;).

What differences? I haven't seen anything that demonstrates that lusers 
isn't just a different wording for "losers". The quote you provided:

"It went from saying "x users logged in" when you pressed C-z, to saying 
"x losers logged in".  Some users, apparently, didn't like being called 
losers, so the compromise "lusers" was reached.  It stuck."

Appears to demonstrate that the spelling was changed from "losers" to 
"lusers" as a result of a compromise. Perhaps I'm missing something but 
the compromise appears to be just as insulting as the original. The only 
thing that changed was the spelling. What compromised was reached? Why 
not change it back to "x users"? Am I missing something?

Josh

On Friday 21 November 2003 11:13 pm in comp.sys.sun.admin Josh McKee wrote:


> Appears to demonstrate that the spelling was changed from "losers" to
> "lusers" as a result of a compromise. Perhaps I'm missing something but
> the compromise appears to be just as insulting as the original. The only
> thing that changed was the spelling. What compromised was reached? Why
> not change it back to "x users"? Am I missing something?

A severe lack of understanding of the Unix sysadmin cultural heritage,
combined with a total lack of any sense of humour.
This is comp.sys.sun.admin which is intended as a forum for Solaris
sysadmins. Maybe you do not belong to this illustrious profession
of caffiene addicted denizens of the data centre ?.

Perhaps a google of the BOFH archives would enlighten you, but a
sense of humour is a prerequisite.

-- 
My real address is crn (at) netunix (dot) com
WARNING all messages containing attachments or html will be silently
deleted. Send only plain text.

In article <1835134.4js5qhy2Q3@callisto>,
 Chris Newport <me@see-my-sig.invalid> wrote:

> On Friday 21 November 2003 11:13 pm in comp.sys.sun.admin Josh McKee wrote:
> 
> 
> > Appears to demonstrate that the spelling was changed from "losers" to
> > "lusers" as a result of a compromise. Perhaps I'm missing something but
> > the compromise appears to be just as insulting as the original. The only
> > thing that changed was the spelling. What compromised was reached? Why
> > not change it back to "x users"? Am I missing something?
> 
> A severe lack of understanding of the Unix sysadmin cultural heritage,
> combined with a total lack of any sense of humour.
> This is comp.sys.sun.admin which is intended as a forum for Solaris
> sysadmins. Maybe you do not belong to this illustrious profession
> of caffiene addicted denizens of the data centre ?.
> 
> Perhaps a google of the BOFH archives would enlighten you, but a
> sense of humour is a prerequisite.

A usual your inability to answer the question isn't surprising.

Josh

Josh McKee wrote:

> In article <bplro6$a6p$1@news.ox.ac.uk>,
>  Graham Lee <graham.lee@wadham.oxford.ac.invalid.uk> wrote:
> 
>> Josh McKee wrote:
>> 
>> > In article <LKWdnW8V9NemYCGiRVn-ig@speakeasy.net>,
>> >  Thomas H Jones II <ferric@xanthia.com> wrote:
>> > 
>> >> Besides, the term goes back to a time when it was common to call
>> >> functional design problems "lossage" or to characterize something
>> >> that was in need of repair as "losing". You take a user who has
>> >> a problem (i.e., something that needs fixed), you have a "losing
>> >> user", or l'user or "luser". "Luser" is not fundamentally the same
>> >> as a "loser". However, a "luser" transitions to a "loser" by having
>> >> way too many problems and an overall disposition of helplessness.
>> > 
>> > Spin it any way that you want. The word "luser" appears to be
>> > synonymous with "loser".
>> > 
>> 
>> Only if you deliberately fail to take into account all of the
>> explanations
>> about their differences given in this froup as well as elsewhere.  Which
>> you're perfectly entitled to do of course; just don't expect everyone
>> else to do it too ;).
> 
> What differences? I haven't seen anything that demonstrates that lusers
> isn't just a different wording for "losers". The quote you provided:
> 
> "It went from saying "x users logged in" when you pressed C-z, to saying
> "x losers logged in".  Some users, apparently, didn't like being called
> losers, so the compromise "lusers" was reached.  It stuck."
> 
> Appears to demonstrate that the spelling was changed from "losers" to
> "lusers" as a result of a compromise. Perhaps I'm missing something but
> the compromise appears to be just as insulting as the original. The only
> thing that changed was the spelling. What compromised was reached? Why
> not change it back to "x users"? Am I missing something?
> 
> Josh

That wasn't the only definition given, was it?  What about the one you
quoted above?  BTW, I would tend to agree with Chris Newport here: you may
not think that the standard BOFH humour is particularly funny, but it's
standard in a de facto way.  So many of us will be using it.  I'm afraid
that funny or not it will need some getting used to :)

Graham.

-- 
Graham Lee
Wadham College
Oxford

On Friday 21 November 2003 11:57 pm in comp.sys.sun.admin Josh McKee wrote:

> In article <1835134.4js5qhy2Q3@callisto>,
>  Chris Newport <me@see-my-sig.invalid> wrote:
> 
>> On Friday 21 November 2003 11:13 pm in comp.sys.sun.admin Josh McKee
>> wrote:
>> 
>> 
>> > Appears to demonstrate that the spelling was changed from "losers" to
>> > "lusers" as a result of a compromise. Perhaps I'm missing something but
>> > the compromise appears to be just as insulting as the original. The
>> > only thing that changed was the spelling. What compromised was reached?
>> > Why not change it back to "x users"? Am I missing something?
>> 
>> A severe lack of understanding of the Unix sysadmin cultural heritage,
>> combined with a total lack of any sense of humour.
>> This is comp.sys.sun.admin which is intended as a forum for Solaris
>> sysadmins. Maybe you do not belong to this illustrious profession
>> of caffiene addicted denizens of the data centre ?.
>> 
>> Perhaps a google of the BOFH archives would enlighten you, but a
>> sense of humour is a prerequisite.
> 
> A usual your inability to answer the question isn't surprising.

You just dont get it, do you?

Every question has been answered.
Maybe you lack the wit to understand.
Maybe you are trolling.
There are none so deaf as those who will not hear.

-- 
My real address is crn (at) netunix (dot) com
WARNING all messages containing attachments or html will be silently
deleted. Send only plain text.

"Chris Newport" <me@see-my-sig.invalid> wrote in message
news:1396902.SmtKaRSlRj@callisto...
> On Friday 21 November 2003 11:57 pm in comp.sys.sun.admin Josh McKee
wrote:
> Every question has been answered.
> Maybe you lack the wit to understand.
> Maybe you are trolling.
> There are none so deaf as those who will not hear.
>


Chris I agree with you .
I tried also to answer this Josh ( or Joshua ), but he is still complaining
about the word loser / luser etc ....
Furthermore , he forget that M$ has a big part on the Market because all PC
are automatically deliverd with Windows on it by every SuperMarket in the
World ( but is not delivered by the reseller , you never knows the customer
could help himself to recover ).
He has also no sense of humor ( read the manpage or understand what is the
word best ) .

It is a pitty that we are know all writing mails within a group which is
helpfull for Unix guys .

I decided to never read any mail from Josh ....
Pidbel

In article <1396902.SmtKaRSlRj@callisto>,
 Chris Newport <me@see-my-sig.invalid> wrote:

> On Friday 21 November 2003 11:57 pm in comp.sys.sun.admin Josh McKee wrote:
> 
> > In article <1835134.4js5qhy2Q3@callisto>,
> >  Chris Newport <me@see-my-sig.invalid> wrote:
> > 
> >> On Friday 21 November 2003 11:13 pm in comp.sys.sun.admin Josh McKee
> >> wrote:
> >> 
> >> 
> >> > Appears to demonstrate that the spelling was changed from "losers" to
> >> > "lusers" as a result of a compromise. Perhaps I'm missing something but
> >> > the compromise appears to be just as insulting as the original. The
> >> > only thing that changed was the spelling. What compromised was reached?
> >> > Why not change it back to "x users"? Am I missing something?
> >> 
> >> A severe lack of understanding of the Unix sysadmin cultural heritage,
> >> combined with a total lack of any sense of humour.
> >> This is comp.sys.sun.admin which is intended as a forum for Solaris
> >> sysadmins. Maybe you do not belong to this illustrious profession
> >> of caffiene addicted denizens of the data centre ?.
> >> 
> >> Perhaps a google of the BOFH archives would enlighten you, but a
> >> sense of humour is a prerequisite.
> > 
> > A usual your inability to answer the question isn't surprising.
> 
> You just dont get it, do you?

Apparently I don't.

> Every question has been answered.

Where? As far as I can tell you have avoided addressing my questions.

> Maybe you lack the wit to understand.

I doubt that's it.

> Maybe you are trolling.
> There are none so deaf as those who will not hear.

I'm listening...there's just nothing to hear.

Josh

In article <bpn875$ftv$1@news-reader4.wanadoo.fr>,
 "pidbel" <pidbel@pidbel.pidbel.sdf> wrote:

> "Chris Newport" <me@see-my-sig.invalid> wrote in message
> news:1396902.SmtKaRSlRj@callisto...
> > On Friday 21 November 2003 11:57 pm in comp.sys.sun.admin Josh McKee
> wrote:
> > Every question has been answered.
> > Maybe you lack the wit to understand.
> > Maybe you are trolling.
> > There are none so deaf as those who will not hear.
> >
> 
> 
> Chris I agree with you .
> I tried also to answer this Josh ( or Joshua ), but he is still complaining
> about the word loser / luser etc ....

You bet...I haven't been shown a reason why Chris is calling his users 
losers.

> Furthermore , he forget that M$ has a big part on the Market because all PC
> are automatically deliverd with Windows on it by every SuperMarket in the
> World ( but is not delivered by the reseller , you never knows the customer
> could help himself to recover ).

Not at all. But if I were to sit down an average person in front of a 
UNIX system and a Windows system I would be willing to bet that the 
majority of them would prefer the Windows system due to it's much more 
user friendly interface. You can bury your head in the sand if you'd 
like but the reality is that an average user values GUI type interfaces 
like Windows over the technical merits of UNIX. That reality. The only 
exception may be Mac OS X.

> He has also no sense of humor ( read the manpage or understand what is the
> word best ) .

I've got a great sense of humor. But Chris' response didn't appear to 
suggest humor. I asked why he called them lusers and what response did I 
get? It certainly wasn't "it was humor and here's why..." Perhaps if 
Chris would be so good as to answer a question when asked he may not be 
challenged.

> It is a pitty that we are know all writing mails within a group which is
> helpfull for Unix guys .
> 
> I decided to never read any mail from Josh ....

That's a shame...you'll probably miss out on a lot of useful information 
that I'll probably contribute in the future. Some of you guys are really 
thinned skinned. I don't see what I've written as insulting towards 
anyone yet you're acting as if I had. Your behavior is more of children 
throwing a tantrum.

Josh

Josh McKee <jtmckee@rm-bogus-ac.net> writes:

[18 lines snipped]

>You bet...I haven't been shown a reason why Chris is calling his users 
>losers.

You, however, have shown plenty of reasons why you're a prime killfile
candidate.

Bye.

-- 
       "The road to Paradise is through Intercourse."
The uk.transport FAQ; http://www.huge.org.uk/transport/FAQ.html
        [email me at huge [at] huge [dot] org [dot] uk]

Josh McKee wrote:

> In article <bpn875$ftv$1@news-reader4.wanadoo.fr>,
>  "pidbel" <pidbel@pidbel.pidbel.sdf> wrote:
> 
>> "Chris Newport" <me@see-my-sig.invalid> wrote in message
>> news:1396902.SmtKaRSlRj@callisto...
>> > On Friday 21 November 2003 11:57 pm in comp.sys.sun.admin Josh McKee
>> wrote:
>> > Every question has been answered.
>> > Maybe you lack the wit to understand.
>> > Maybe you are trolling.
>> > There are none so deaf as those who will not hear.
>> >
>> 
>> 
>> Chris I agree with you .
>> I tried also to answer this Josh ( or Joshua ), but he is still
>> complaining about the word loser / luser etc ....
> 
> You bet...I haven't been shown a reason why Chris is calling his users
> losers.

He doesn't; he refers to users as lusers.

> 
>> Furthermore , he forget that M$ has a big part on the Market because all
>> PC are automatically deliverd with Windows on it by every SuperMarket in
>> the World ( but is not delivered by the reseller , you never knows the
>> customer could help himself to recover ).
> 
> Not at all. But if I were to sit down an average person in front of a
> UNIX system and a Windows system I would be willing to bet that the
> majority of them would prefer the Windows system due to it's much more
> user friendly interface.

[first point: due to it *is* much more user friendly interface?  Danger Will
Robinson, spurious apostrophe overload!]

Except that, as you semi-point out later, NEXTSTEP->Mac OS X have great user
interfaces.  KDE and GNOME are 'not bad', and most people I've exposed them
to think they're perfectly usable.  OK, so CDE is a bit rubbish, but at
least it isn't BrokenWindows ;).

Interestingly, a large number of Windows users have come up to me asking for
alternatives, because their computers are unstable, which leads to annoying
freezes and loss of data.  People who use work machines[*] respect their
admin more if the machines stay running, so that they can get their jobs
done.  UNIX helps my colleagues to love me :).  ANd 'Joe Avergae'-type
users are realising that there's more to life than animated paper clips.

> You can bury your head in the sand if you'd
> like but the reality is that an average user values GUI type interfaces
> like Windows over the technical merits of UNIX. That reality. The only
> exception may be Mac OS X.

The average user in my experience seems to value both user-friendliness and
functionality.  Mac OS X is only an 'exception' in that it is the only
system on the market to satisfactorily deliver both.  That is, unless you
count some GNU/Linux distro that comes bundled with a decent UI and a
stable system, but I don't because of the 'KDOME' problem[**].

> 
>> He has also no sense of humor ( read the manpage or understand what is
>> the word best ) .
> 
> I've got a great sense of humor.

I think it's a little presumptuous to assume that you're allowed to decide
that ;).

> But Chris' response didn't appear to
> suggest humor.

Really?  The fact that the word 'luser' is a pun (or play on words) seemed
to imply humour.  And we're UNIX people around here; "IMPLICIT NONE" is
definitely not an option :).  Chris' reply to your post definitely did have
a reek of 'tongue-in-cheek' about it; I'd say that humour was very strongly
suggested.

> I asked why he called them lusers and what response did I
> get? It certainly wasn't "it was humor and here's why..." Perhaps if
> Chris would be so good as to answer a question when asked he may not be
> challenged.

See above.

> 
>> It is a pitty that we are know all writing mails within a group which is
>> helpfull for Unix guys .
>> 
>> I decided to never read any mail from Josh ....
> 
> That's a shame...you'll probably miss out on a lot of useful information
> that I'll probably contribute in the future.

[paragraph separated for emphasis]

> Some of you guys are really
> thinned skinned. I don't see what I've written as insulting towards
> anyone yet you're acting as if I had. Your behavior is more of children
> throwing a tantrum.

I hope the irony here was intentional :).

[*]'work workstations' seemed a little redundant.
[**]For instance, if you're running KDE and load a GNOME app, it *really*
looks different.  Even on Red Hat, before anyone tries that card :P. 
Everyone should use GNUstep, that'd solve the problem</holywar>.
-- 
Graham Lee
Wadham College
Oxford

On Saturday 22 November 2003 3:46 pm in comp.sys.sun.admin Josh McKee wrote:


>> Chris I agree with you .
>> I tried also to answer this Josh ( or Joshua ), but he is still
>> complaining about the word loser / luser etc ....
> 
> You bet...I haven't been shown a reason why Chris is calling his users
> losers.

Oh dear, deliberate false accusations ... time to shout....

NO I DID NOT, DUMKOMPF.

Please remove your head from your arse and read what I wrote, not
what you misread.

Luser != loser.
Luser == lame user
Users wanted long login names
QED


-- 
My real address is crn (at) netunix (dot) com
WARNING all messages containing attachments or html will be silently
deleted. Send only plain text.

>Josh McKee wrote:

[30 lines snipped]

>> Not at all. But if I were to sit down an average person in front of a
>> UNIX system and a Windows system I would be willing to bet that the
>> majority of them would prefer the Windows system due to it's much more
>> user friendly interface.

Wrong. It's not more user friendly, it's merely more familiar.

Until very recently, I'd never used Windows in anger, having started on
windowing interfaces with the Xerox Star and Interlisp and moving from
there to Unix systems, mostly Sun. To this day, I find Windows
confusing and frustrating, sometimes to the point of incoherent rage.

>> You can bury your head in the sand if you'd
>> like but the reality is that an average user values GUI type interfaces
>> like Windows over the technical merits of UNIX. That reality.

Not "reality". Ignorance.


-- 
       "The road to Paradise is through Intercourse."
The uk.transport FAQ; http://www.huge.org.uk/transport/FAQ.html
        [email me at huge [at] huge [dot] org [dot] uk]

On Sat, 22 Nov 2003 17:13:50 +0000,
Graham Lee <graham.lee@wadham.oxford.ac.invalid.uk>, in
<bpo5gl$4de$1@news.ox.ac.uk> wrote:

+> And we're UNIX people around here; "IMPLICIT NONE" is definitely
+> not an option :).

Here now. As an old f77 hand, I resent that implication.

:-)

James
-- 
Consulting Minister for Consultants, DNRC
I can please only one person per day. Today is not your day. Tomorrow
isn't looking good, either.
I am BOFH. Resistance is futile. Your network will be assimilated.

In article <jtmckee-87DA11.08463822112003@netnews.attbi.com>,
Josh McKee  <jtmckee@rm-bogus-ac.net> wrote:
>
>Not at all. But if I were to sit down an average person in front of a 
>UNIX system and a Windows system I would be willing to bet that the 
>majority of them would prefer the Windows system due to it's much more 
>user friendly interface. You can bury your head in the sand if you'd 
>like but the reality is that an average user values GUI type interfaces 
>like Windows over the technical merits of UNIX. That reality. The only 
>exception may be Mac OS X.

There are no average people. The computer world seems to have two kinds,
distinctly above average and distinctly below.

I've constantly heard claims that windows is more user friendly but never
any data to really prove it. Once windows 95 got close to the clear lead
the Mac GUI had, all the OSes became more or less the same GUIs engines. 
It is driving different cars, sitting in front of a new car requires
some changes, but you get use to it.

To most business people all the computers are the same. Heck many people
never use anything but a browser.

I'm fact, I don't think there are many people around that picked their
computer system. We all went to work somewhere, and use whatever computers
were there. I know lots of ms people that advised parents to buy the
computer like the one at work because kids could learn new systems faster.
I think they did even acknowledge that the mac was a better choice for the
kids.
-- 
http://www.math.fsu.edu/~bellenot
bellenot <At/> math.fsu.edu 
+1.850.644.7189 (4053fax)

On 22 Nov 2003 22:32:27 GMT, bellenot@math.fsu.edu (Steve Bellenot)
wrote:

>In article <jtmckee-87DA11.08463822112003@netnews.attbi.com>,
>Josh McKee  <jtmckee@rm-bogus-ac.net> wrote:
>>
>>Not at all. But if I were to sit down an average person in front of a 
>>UNIX system and a Windows system I would be willing to bet that the 
>>majority of them would prefer the Windows system due to it's much more 
>>user friendly interface. You can bury your head in the sand if you'd 
>>like but the reality is that an average user values GUI type interfaces 
>>like Windows over the technical merits of UNIX. That reality. The only 
>>exception may be Mac OS X.
>
>There are no average people. The computer world seems to have two kinds,
>distinctly above average and distinctly below.

Only if you want to use the mathematical definition of the word. The
average person will understand that the mathematical definition is not
the appropriate one.

>I've constantly heard claims that windows is more user friendly but never
>any data to really prove it. Once windows 95 got close to the clear lead
>the Mac GUI had, all the OSes became more or less the same GUIs engines. 
>It is driving different cars, sitting in front of a new car requires
>some changes, but you get use to it.

They require some changes but not drastic changes. Give an average
user a Windows system and a command line Linux system and I'll bet
that they'll prefer Windows.

>To most business people all the computers are the same. Heck many people
>never use anything but a browser.

I think that they're becoming more the same...they're becoming more
like Windows and Mac OS.

>I'm fact, I don't think there are many people around that picked their
>computer system. We all went to work somewhere, and use whatever computers
>were there.

I bet that given a choice between Windows and CLI based UNIX they
would prefer Windows regardless of what they use at work. I think that
MS-DOS demonstrated this quite well. People used MS-DOS at work
because that's what their employers gave them. When Windows hit the
market the preference was towards Windows. And that's
understandable...it offered a much better UI.

What I find surprising is that you're even questioning this when there
is so much empirical evidence to support it.

>I know lots of ms people that advised parents to buy the
>computer like the one at work because kids could learn new systems faster.
>I think they did even acknowledge that the mac was a better choice for the
>kids.

And why was the Mac a better choice? Could it be that it had a better
UI?

Josh

On Sat, 22 Nov 2003, Josh McKee wrote:

> I bet that given a choice between Windows and CLI based UNIX they

Oh come on!  What desktop UNIX (or UNIX-like) system these
days DOESn'T have a GUI?

> would prefer Windows regardless of what they use at work. I think that

I submit that most Windoze user use it because they are
not even aware that there's a choice.

-- 
Rich Teer, SCNA, SCSA

President,
Rite Online Inc.

Voice: +1 (250) 979-1638
URL: http://www.rite-online.net

In article <Pine.SOL.4.58.0311221726510.507@zaphod.rite-group.com>,
 Rich Teer <rich.teer@rite-group.com> wrote:

> On Sat, 22 Nov 2003, Josh McKee wrote:
> 
> > I bet that given a choice between Windows and CLI based UNIX they
> 
> Oh come on!  What desktop UNIX (or UNIX-like) system these
> days DOESn'T have a GUI?

I don't know. Why?

> > would prefer Windows regardless of what they use at work. I think that
> 
> I submit that most Windoze user use it because they are
> not even aware that there's a choice.

I would submit that Windows users, given a choice between a CLI based 
UNIX system and Windows, would choose Windows.

Josh

BTW - Is your use of the word "Windoze" some hidden humor like "lusers"?

On Sun, 23 Nov 2003 02:46:06 GMT,
Josh McKee <jtmckee@rm-bogus-ac.net>, in
<jtmckee-B6E50F.19460522112003@netnews.attbi.com> wrote:

+> I would submit that Windows users, given a choice between a CLI based 
+> UNIX system and Windows, would choose Windows.

I submit that you are an idiot.

James
-- 
Consulting Minister for Consultants, DNRC
I can please only one person per day. Today is not your day. Tomorrow
isn't looking good, either.
I am BOFH. Resistance is futile. Your network will be assimilated.

In article <slrnbs08vu.58u.sy_nttvr@gurcragntba.pbz>,
 sy_nttvr@gurcragntba.pbz (I R A Darth Aggie) wrote:

> On Sun, 23 Nov 2003 02:46:06 GMT,
> Josh McKee <jtmckee@rm-bogus-ac.net>, in
> <jtmckee-B6E50F.19460522112003@netnews.attbi.com> wrote:
> 
> +> I would submit that Windows users, given a choice between a CLI based 
> +> UNIX system and Windows, would choose Windows.
> 
> I submit that you are an idiot.

Why is that? Because I have a differnce of opinion than you?

Josh

On Sun, 23 Nov 2003, Josh McKee wrote:

> In article <Pine.SOL.4.58.0311221726510.507@zaphod.rite-group.com>,
>  Rich Teer <rich.teer@rite-group.com> wrote:
>
> > On Sat, 22 Nov 2003, Josh McKee wrote:
> >
> > > I bet that given a choice between Windows and CLI based UNIX they
> >
> > Oh come on!  What desktop UNIX (or UNIX-like) system these
> > days DOESn'T have a GUI?
>
> I don't know. Why?

That's my point.  I don't think there is a modern desktop
UNIX (or UNIX-like) system that is soley command line based.
Therefore, you're comparing apples to oranges.

> I would submit that Windows users, given a choice between a CLI based
> UNIX system and Windows, would choose Windows.

Of course the Windoze users would prefer Windoze - that's what
they're used to!  Even that blindingly obvious point aside, I
would agree that users who have never used a computer are more
likely to prefer Windoze over a CLI UNIX.  However, I think the
results would be far more evenly distributed when Windoze is
compared against a modern UNIX GUI.

> BTW - Is your use of the word "Windoze" some hidden humor like "lusers"?

No, I mean derogatorially.  "lusers" may or may not be derogatory,
depending on context.

-- 
Rich Teer, SCNA, SCSA

President,
Rite Online Inc.

Voice: +1 (250) 979-1638
URL: http://www.rite-online.net

In article <nlqvrvc539esjsjiplarrr2au34hl23a38@4ax.com>,
Josh McKee  <jtmckee@rm-bogus-ac.net> wrote:
>They require some changes but not drastic changes. Give an average
>user a Windows system and a command line Linux system and I'll bet
>that they'll prefer Windows.

Your equation of Unix = command line only hasn't been true since at
least the mid to late 80's. The fact is that the command line needs to
be never used as lots of Mac OS X users know. Heck, the only reason
many Mac users know that there is unix under the hood is because Apple
knows that Unix is thought of as being more stable than any windows
system, and uses the fact in their ads.
-- 
http://www.math.fsu.edu/~bellenot
bellenot <At/> math.fsu.edu 
+1.850.644.7189 (4053fax)

I R A Darth Aggie wrote:
> On Sat, 22 Nov 2003 17:13:50 +0000,
> Graham Lee <graham.lee@wadham.oxford.ac.invalid.uk>, in
> <bpo5gl$4de$1@news.ox.ac.uk> wrote:
> 
> +> And we're UNIX people around here; "IMPLICIT NONE" is definitely
> +> not an option :).
> 
> Here now. As an old f77 hand, I resent that implication.
> 
> :-)
> 
> James

COBOL (Completely Over and Beyond Obvious Logic) would punch f77 in the 
eye ;-)

In article <bpq2j7$nbm$1@news.fsu.edu>,
 bellenot@math.fsu.edu (Steve Bellenot) wrote:

> In article <nlqvrvc539esjsjiplarrr2au34hl23a38@4ax.com>,
> Josh McKee  <jtmckee@rm-bogus-ac.net> wrote:
> >They require some changes but not drastic changes. Give an average
> >user a Windows system and a command line Linux system and I'll bet
> >that they'll prefer Windows.
> 
> Your equation of Unix = command line only hasn't been true since at
> least the mid to late 80's.

I am aware of this. Your point?

Josh

Approximately 11/22/03 19:05, I R A Darth Aggie uttered for posterity:

> On Sun, 23 Nov 2003 02:46:06 GMT,
> Josh McKee <jtmckee@rm-bogus-ac.net>, in
> <jtmckee-B6E50F.19460522112003@netnews.attbi.com> wrote:
> 
> +> I would submit that Windows users, given a choice between a CLI based 
> +> UNIX system and Windows, would choose Windows.
> 
> I submit that you are an idiot.
> 
   I submit both of you are correct.
-- 
Still a Raiders fan, but no longer sure why.

On Sun, 23 Nov 2003 19:43:54 GMT, Lon Stowell
<LonDot.Stowell@ComcastPeriod.Net> wrote:

>Approximately 11/22/03 19:05, I R A Darth Aggie uttered for posterity:
>
>> On Sun, 23 Nov 2003 02:46:06 GMT,
>> Josh McKee <jtmckee@rm-bogus-ac.net>, in
>> <jtmckee-B6E50F.19460522112003@netnews.attbi.com> wrote:
>> 
>> +> I would submit that Windows users, given a choice between a CLI based 
>> +> UNIX system and Windows, would choose Windows.
>> 
>> I submit that you are an idiot.
>> 
>   I submit both of you are correct.

That doesn't follow. How can I be correct and an idiot?

Josh

On Sun, 23 Nov 2003 04:33:10 GMT, Rich Teer <rich.teer@rite-group.com>
wrote:

>On Sun, 23 Nov 2003, Josh McKee wrote:
>
>> In article <Pine.SOL.4.58.0311221726510.507@zaphod.rite-group.com>,
>>  Rich Teer <rich.teer@rite-group.com> wrote:
>>
>> > On Sat, 22 Nov 2003, Josh McKee wrote:
>> >
>> > > I bet that given a choice between Windows and CLI based UNIX they
>> >
>> > Oh come on!  What desktop UNIX (or UNIX-like) system these
>> > days DOESn'T have a GUI?
>>
>> I don't know. Why?
>
>That's my point.  I don't think there is a modern desktop
>UNIX (or UNIX-like) system that is soley command line based.

Did I claim otherwise?

>Therefore, you're comparing apples to oranges.

How so? The stability of a UNIX system is the same regardless of if a
CLI or GUI interface is used is it not?

>> I would submit that Windows users, given a choice between a CLI based
>> UNIX system and Windows, would choose Windows.
>
>Of course the Windoze users would prefer Windoze - that's what
>they're used to!  Even that blindingly obvious point aside, I
>would agree that users who have never used a computer are more
>likely to prefer Windoze over a CLI UNIX.  However, I think the
>results would be far more evenly distributed when Windoze is
>compared against a modern UNIX GUI.

Don't look now Rich but you've just argued my point. Recall that I
claimed that an easier to use interface is more important to the
average user than a stable system that has a more difficult to use
user interface.

>> BTW - Is your use of the word "Windoze" some hidden humor like "lusers"?
>
>No, I mean derogatorially.  "lusers" may or may not be derogatory,
>depending on context.

Exactly...so it appears that I am completely in the right to assume
that Chris was using the term in a derogatory manner.

Josh

Josh McKee <jtmckee@rm-bogus-ac.net> wrote:
> The problem I have is Chris' portrayal of the user as a loser because
> they wish to use something more descriptive. 

	No one commented on this, but I think this is the philosophical
crux of the matter.  The username is not supposed to be descriptive, it's
supposed to be a unique identifier.  If you want descriptive, that's what
the GECOS field is for.  

"As a general rule, I am adamantly opposed to using full names as
e-mail addresses, since they are not in any sense unique.  For example,
the Unix software-development community has two Andy Tannenbaums,
at least two well-known Peter Deutsches, and at one time Bell Labs
had two Stephen R. Bournes with offices along the same hallway."
	-- Eric Allman

In article <jtmckee-13DE6F.10285723112003@netnews.attbi.com>,
	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
> In article <bpq2j7$nbm$1@news.fsu.edu>,
>  bellenot@math.fsu.edu (Steve Bellenot) wrote:
>> In article <nlqvrvc539esjsjiplarrr2au34hl23a38@4ax.com>,
>> Josh McKee  <jtmckee@rm-bogus-ac.net> wrote:
>> >They require some changes but not drastic changes. Give an average
>> >user a Windows system and a command line Linux system and I'll bet
>> >that they'll prefer Windows.
>> Your equation of Unix = command line only hasn't been true since at
>> least the mid to late 80's.
> I am aware of this. Your point?

What's yours?

"But I came here to argue...!
"O this is abuse.."

In article <bptapg$3l8$2@reader2.panix.com>,
 Jeremiah DeWitt Weiner <jdw@panix.com> wrote:

> Josh McKee <jtmckee@rm-bogus-ac.net> wrote:
> > The problem I have is Chris' portrayal of the user as a loser because
> > they wish to use something more descriptive. 
> 
> 	No one commented on this, but I think this is the philosophical
> crux of the matter.  The username is not supposed to be descriptive,

Why not? Especially if that's what the users want.

> it's supposed to be a unique identifier.  If you want descriptive, that's what
> the GECOS field is for.  

Users don't log in with their GECOS field.

> "As a general rule, I am adamantly opposed to using full names as
> e-mail addresses, since they are not in any sense unique.  For example,
> the Unix software-development community has two Andy Tannenbaums,
> at least two well-known Peter Deutsches, and at one time Bell Labs
> had two Stephen R. Bournes with offices along the same hallway."
> 	-- Eric Allman

Eric should learn to listen to the user and not force his opinion on 
them. The user is the customer. We shouldn't lose sight of that. Phone 
books listen numbers by full names. Why should e-mail addresses be any 
different? If there's a discrepency that's easily solved by appending a 
numerical count to someones name.

Josh

In article <bSqwb.14855$f7.727553@localhost>, gerryt@gtconnect.net () 
wrote:

> In article <jtmckee-13DE6F.10285723112003@netnews.attbi.com>,
> 	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
> > In article <bpq2j7$nbm$1@news.fsu.edu>,
> >  bellenot@math.fsu.edu (Steve Bellenot) wrote:
> >> In article <nlqvrvc539esjsjiplarrr2au34hl23a38@4ax.com>,
> >> Josh McKee  <jtmckee@rm-bogus-ac.net> wrote:
> >> >They require some changes but not drastic changes. Give an average
> >> >user a Windows system and a command line Linux system and I'll bet
> >> >that they'll prefer Windows.
> >> Your equation of Unix = command line only hasn't been true since at
> >> least the mid to late 80's.
> > I am aware of this. Your point?
> 
> What's yours?

That the user interface is more important to the average user than the 
technical merits of UNIX. And a system that fits the users requirements 
is going to be preferred by them. Thus UNIX should try to adapt to the 
users wants instead of forcing the user to adapt to the systems 
requirements. It's a realls simple concept: Computers should accommodate 
us, not the other way around. Do you disagree?

Josh

> "But I came here to argue...!
> "O this is abuse.."

Josh McKee wrote:

> In article <bptapg$3l8$2@reader2.panix.com>,
>  Jeremiah DeWitt Weiner <jdw@panix.com> wrote:
> 
>> Josh McKee <jtmckee@rm-bogus-ac.net> wrote:
>> > The problem I have is Chris' portrayal of the user as a loser because
>> > they wish to use something more descriptive.
>> 
>> No one commented on this, but I think this is the philosophical
>> crux of the matter.  The username is not supposed to be descriptive,
> 
> Why not? Especially if that's what the users want.
> 
>> it's supposed to be a unique identifier.  If you want descriptive, that's
>> what the GECOS field is for.
> 
> Users don't log in with their GECOS field.
> 

Observe any of the xdm logins: they (apart from the one on Slowlaris,
admittedly) display the GECOS information along with the username.  So that
as long as you boot into runlevel 5, lusers *can* log in with their GECOS
field.  The username is merely used internally to match the passwd and
shadow information.
-- 
Graham Lee
Wadham College
Oxford

On Sun, 23 Nov 2003 03:31:08 GMT,
Josh McKee <jtmckee@rm-bogus-ac.net>, in
<jtmckee-D420A0.20310822112003@netnews.attbi.com> wrote:
+> In article <slrnbs08vu.58u.sy_nttvr@gurcragntba.pbz>,
+>  sy_nttvr@gurcragntba.pbz (I R A Darth Aggie) wrote:
+> 
+> > On Sun, 23 Nov 2003 02:46:06 GMT,
+> > Josh McKee <jtmckee@rm-bogus-ac.net>, in
+> > <jtmckee-B6E50F.19460522112003@netnews.attbi.com> wrote:
+> > 
+> > +> I would submit that Windows users, given a choice between a CLI based 
+> > +> UNIX system and Windows, would choose Windows.
+> > 
+> > I submit that you are an idiot.
+> 
+> Why is that? Because I have a differnce of opinion than you?

No, because I have a different opinion than you. You have single
minded determination that *you*are*right*, and once you've come to
that conclusion then there is naught anyone can do to sway you.

Further, you quibble over the definition of "average" with a
mathemagician?

James
-- 
Consulting Minister for Consultants, DNRC
I can please only one person per day. Today is not your day. Tomorrow
isn't looking good, either.
I am BOFH. Resistance is futile. Your network will be assimilated.

On Sun, 23 Nov 2003 13:10:14 +0000,
Beardy <beardy@beardy.net>, in
<im2wb.12010$lm1.86751@wards.force9.net> wrote:
+> I R A Darth Aggie wrote:
+> > On Sat, 22 Nov 2003 17:13:50 +0000,
+> > Graham Lee <graham.lee@wadham.oxford.ac.invalid.uk>, in
+> > <bpo5gl$4de$1@news.ox.ac.uk> wrote:
+> > 
+> > +> And we're UNIX people around here; "IMPLICIT NONE" is definitely
+> > +> not an option :).
+> > 
+> > Here now. As an old f77 hand, I resent that implication.
+> > 
+> > :-)
+> > 
+> > James
+> 
+> COBOL (Completely Over and Beyond Obvious Logic) would punch f77 in the 
+> eye ;-)

Don't make me come over there and snap you into next week using your
suspenders! 

James
-- 
Consulting Minister for Consultants, DNRC
I can please only one person per day. Today is not your day. Tomorrow
isn't looking good, either.
I am BOFH. Resistance is futile. Your network will be assimilated.

On Tue, 25 Nov 2003 00:42:37 GMT,
Josh McKee <jtmckee@rm-bogus-ac.net>, in
<jtmckee-090C55.17423524112003@netnews.attbi.com> wrote:
+> In article <bptapg$3l8$2@reader2.panix.com>,
+>  Jeremiah DeWitt Weiner <jdw@panix.com> wrote:
+> 
+> > Josh McKee <jtmckee@rm-bogus-ac.net> wrote:
+> > > The problem I have is Chris' portrayal of the user as a loser because
+> > > they wish to use something more descriptive. 
+> > 
+> > 	No one commented on this, but I think this is the philosophical
+> > crux of the matter.  The username is not supposed to be descriptive,
+> 
+> Why not? Especially if that's what the users want.
+> 
+> > it's supposed to be a unique identifier.  If you want descriptive, that's what
+> > the GECOS field is for.  
+> 
+> Users don't log in with their GECOS field.
+> 
+> > "As a general rule, I am adamantly opposed to using full names as
+> > e-mail addresses, since they are not in any sense unique.  For example,
+> > the Unix software-development community has two Andy Tannenbaums,
+> > at least two well-known Peter Deutsches, and at one time Bell Labs
+> > had two Stephen R. Bournes with offices along the same hallway."
+> > 	-- Eric Allman
+> 
+> Eric should learn to listen to the user and not force his opinion on 
+> them. The user is the customer. We shouldn't lose sight of that. Phone 
+> books listen numbers by full names. Why should e-mail addresses be any 
+> different? If there's a discrepency that's easily solved by appending a 
+> numerical count to someones name.

What if the user doesn't want that?

James
-- 
Consulting Minister for Consultants, DNRC
I can please only one person per day. Today is not your day. Tomorrow
isn't looking good, either.
I am BOFH. Resistance is futile. Your network will be assimilated.

In article <jtmckee-EB32C2.17452124112003@netnews.attbi.com>,
	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
> In article <bSqwb.14855$f7.727553@localhost>, gerryt@gtconnect.net () 
> wrote:
> 
>> In article <jtmckee-13DE6F.10285723112003@netnews.attbi.com>,
>> 	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
>> > In article <bpq2j7$nbm$1@news.fsu.edu>,
>> >  bellenot@math.fsu.edu (Steve Bellenot) wrote:
>> >> In article <nlqvrvc539esjsjiplarrr2au34hl23a38@4ax.com>,
>> >> Josh McKee  <jtmckee@rm-bogus-ac.net> wrote:
>> >> >They require some changes but not drastic changes. Give an average
>> >> >user a Windows system and a command line Linux system and I'll bet
>> >> >that they'll prefer Windows.
>> >> Your equation of Unix = command line only hasn't been true since at
>> >> least the mid to late 80's.
>> > I am aware of this. Your point?
>> 
>> What's yours?
> 
> That the user interface is more important to the average user than the 
> technical merits of UNIX. And a system that fits the users requirements 
> is going to be preferred by them. Thus UNIX should try to adapt to the 
> users wants instead of forcing the user to adapt to the systems 
> requirements. It's a realls simple concept: Computers should accommodate 
> us, not the other way around. Do you disagree?

I said POINT. If you are talking interface in the second to last
sentence I agree. Interface like editors is a quasi religious
discussion i.e. no one wins that one. Especially lusers : >

>> "But I came here to argue...!
>> "O this is abuse.."

In article <slrnbs70r9.obr.sy_nttvr@gurcragntba.pbz>,
 sy_nttvr@gurcragntba.pbz (I R A Darth Aggie) wrote:

> On Tue, 25 Nov 2003 00:42:37 GMT,
> Josh McKee <jtmckee@rm-bogus-ac.net>, in
> <jtmckee-090C55.17423524112003@netnews.attbi.com> wrote:
> +> In article <bptapg$3l8$2@reader2.panix.com>,
> +>  Jeremiah DeWitt Weiner <jdw@panix.com> wrote:
> +> 
> +> > Josh McKee <jtmckee@rm-bogus-ac.net> wrote:
> +> > > The problem I have is Chris' portrayal of the user as a loser because
> +> > > they wish to use something more descriptive. 
> +> > 
> +> > 	No one commented on this, but I think this is the philosophical
> +> > crux of the matter.  The username is not supposed to be descriptive,
> +> 
> +> Why not? Especially if that's what the users want.
> +> 
> +> > it's supposed to be a unique identifier.  If you want descriptive, 
> that's what
> +> > the GECOS field is for.  
> +> 
> +> Users don't log in with their GECOS field.
> +> 
> +> > "As a general rule, I am adamantly opposed to using full names as
> +> > e-mail addresses, since they are not in any sense unique.  For example,
> +> > the Unix software-development community has two Andy Tannenbaums,
> +> > at least two well-known Peter Deutsches, and at one time Bell Labs
> +> > had two Stephen R. Bournes with offices along the same hallway."
> +> > 	-- Eric Allman
> +> 
> +> Eric should learn to listen to the user and not force his opinion on 
> +> them. The user is the customer. We shouldn't lose sight of that. Phone 
> +> books listen numbers by full names. Why should e-mail addresses be any 
> +> different? If there's a discrepency that's easily solved by appending a 
> +> numerical count to someones name.
> 
> What if the user doesn't want that?

There doesn't appear to be any solution. You'll need to come to an 
agreement on something that's unique.

Josh

In article <bpuj1n$i4k$1@news.ox.ac.uk>,
 Graham Lee <graham.lee@wadham.oxford.ac.invalid.uk> wrote:

> Josh McKee wrote:
> 
> > In article <bptapg$3l8$2@reader2.panix.com>,
> >  Jeremiah DeWitt Weiner <jdw@panix.com> wrote:
> > 
> >> Josh McKee <jtmckee@rm-bogus-ac.net> wrote:
> >> > The problem I have is Chris' portrayal of the user as a loser because
> >> > they wish to use something more descriptive.
> >> 
> >> No one commented on this, but I think this is the philosophical
> >> crux of the matter.  The username is not supposed to be descriptive,
> > 
> > Why not? Especially if that's what the users want.
> > 
> >> it's supposed to be a unique identifier.  If you want descriptive, that's
> >> what the GECOS field is for.
> > 
> > Users don't log in with their GECOS field.
> > 
> 
> Observe any of the xdm logins: they (apart from the one on Slowlaris,
> admittedly) display the GECOS information along with the username.  So that
> as long as you boot into runlevel 5, lusers *can* log in with their GECOS
> field.  The username is merely used internally to match the passwd and
> shadow information.

I was unaware of this...thanks!

Josh

In article <e5kprvg380asilftdn2dpustmml3hkpi43@4ax.com>,
	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
> On Wed, 19 Nov 2003 18:08:12 +0100, "pidbel"
> <pidbel@pidbel.pidbel.sdf> wrote:
> 
>>
>>"Josh McKee" <jtmckee@rm-bogus-ac.net> wrote in message
>>news:mr7nrv8lng5tpgtdgu0immbmog8f07a4d1@4ax.com...
>>> On Tue, 18 Nov 2003 17:33:51 -0800, "Michael Vilain
>>> <vilain@spamcop.net>" wrote:
>>>
>>> And you wonder why Windows is the dominant platform. It may be
>>> technically inferior to UNIX. But it appears to accommodate the user
>>> in the areas that are most important to them.
>>>
>>> Josh
>>
>>Indeed, if you forget the time staring at the blue screen and the daily
>>security patches you have to install , Windows is accommodating the user
>>very well ....
> 
> Yet Windows still commands a huge share of the market. Why is that?
> It's because Windows is giving the user what they want. It should be
> obvious that ease of use is what the users want. Reliability and
> security are taking a back seat to user friendliness. You can present
> your case about the strengths of UNIX and the weaknesses of Windows
> all day long. But in the end the user is going to want the user
> friendliness of Windows over the reliability of UNIX.
>

It's fine to be all worried about what the itty bitty poor baby users
want, but they're like little children; giving them what they want is
_wrong_ if it could hurt them or burn down the house.

One of those strengths of Unix that you seem to disdain in favor of
friendliness is that updates to the operating system almost never break
applications.  Raising the userid limit above 8 characters is one of those
things that could potentially break lots of existing applications.
(if you want to know why it could do that, read the cuserid(3c) man page
and strain your brain a little...)

[...]
> They probably forget their user ID because it's something cryptic
> designed to fit within an 8 character limitation of the system. It
> requires the user to adapt to the system and not the other way around.
> 
>>So what is the case about the 8 character limitations , as some users
>>allready can forget his login of one character :-))))
> 
> But if it were their name how likely do you think that they'd forget
> it? Probably unlikely.

There used to be a suggestion a long time ago in one of the admin manuals
to take their initials and enough letters from the beginning of their last
name to make eight letters; if that wasn't unique, change the last
character to a digit.

A procedure like that produces something that's should be easy enough to
remember.  A procedure is also better than letting users pick their own
names, because the procedure can deal consistently with avoiding conflicts;
no need to take a lot of crap when the name someone wants is already taken.

Lots of operating systems other than Unix have short user names.  At least
until the time it picked up a bunch of Unix compatibility features, I think
at least one of the major IBM mainframe OS's had a seven character limit.
I recall one other OS (long gone now, but that's not why) that had a six
character limit.

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

In article <7nkprv4g4iqm593grkhvq84rven8i9fap5@4ax.com>,
	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
> On Wed, 19 Nov 2003 23:23:00 +0000, Chris Newport
> <me@see-my-sig.invalid> wrote:
> 
>>On Wednesday 19 November 2003 4:54 pm in comp.sys.sun.admin Josh McKee
>>wrote:
>>
>> 
>>>>> Why are you calling them lusers? Shouldn't the system accommodate the
>>>>> user instead of the user accommodating the system?
>>>>
>>>>In this case, the users must accomodate the system, since it's a hard
>>>>limit.  If you think it's cost effective to spend time modifying Solaris
>>>>and all the parts affected by this to accomedate usernames, passwords
>>>>and groups > 8 characters, then by all means have fun.
>>> 
>>> The problem I have is Chris' portrayal of the user as a loser because
>>> they wish to use something more descriptive. I fail to see why that is
>>> an unreasonable expectation and warrants the SA calling their
>>> customers losers. Computers are supposed to work for us, not the other
>>> way around.
>>
>>Oh purleeeese !
>>
>>Are you lysdexic ?.
>>Can you tell the difference between luser and loser ?
> 
> No.
> 
>>The two terms are not the same, although they often overlap.
> 
> Perhaps you'd be willing to explain why loser and luser are not, for
> all intents and purposes, the same? According to dictionary.com they
> appear to be synonymous.

Losers are (barring a miracle) losers forever, whereas some lusers may
earn the right to get the "l" out.

It may not do to tell people to their face that they're probably idiots,
but as a practical matter, if your job is to maintain something that
people use, you need to presume that they're going to _act_ like idiots at
least until such time as specific ones demonstrate otherwise.

If you don't think that's true, just spend 5 minutes watching "Jerry
Springer" or similar shows.  Think of how many pervs and nutcases there
must be out there to let themselves be made fools of for no more than a
free airplane ticket, limo ride, a night in a hotel, and a couple of
meals.  Now think how many more fools there are that watch such garbage to
feel better about themselves 'cause at least they're not _that_ bad.
Individually, people will fall on a range from clueless to very clueful
indeed.  Collectively, although they'll occasionally show a surprising
ability to detect BS, they'll tend to be dumber according to the inverse
of the square of the number of people in question, or at least on some
curve that approaches but never quite reaches the lowest level among
them.  The very survival of dictators also demonstrates that; they'd never
stand a chance if even 20% of their people stood against them, and the
amount of death and misery involved in overthowing them (given that once
you start, you've got to see it through) would be less than that of a few
years of their continuing reign.

[...]
>>The simple solution is to educate the luser and impart the clue
>>that his email address need not be the same as his login name.
> 
> Nothing wrong with that...except that you're classifying them as
> losers instead of users. I disagree with that. Especially when the
> computer should accommodate the user and not the user having to
> accommodate the system. Do you consider that unreasonable?

As compared to breaking backwards compatibility for a bunch of existing
applications, yes.  Being able to run any number of commercial (and
in-house) apps that actually do something useful is more important than
getting the name you want.

It's just imaginable that changes could be made such that user names could
be allowed to be longer as long as they remained unique within the first
eight characters for the sake of existing applications incapable of
handling longer than eight character user names; that could perhaps be
done without breaking backwards compatibility.  There are some reasons
to believe that things are slowly heading in that direction.  But AFAIK
they aren't there yet, so erring on the safe side is only responsible.

There are always constraints, and trade-offs.  People don't take the
time to repeatedly say "you can't have that" just to be mean (they
may do it once or twice to be mean, but it gets old too quickly to
keep that up for long).  They say it because the harm outweighs the
benefits, at least until such time as a consensus is found on how to
achieve a reasonable approximation of what's so often requested while
still honoring functional considerations like backwards compatibility.

Microsoft, being monolithic, can break whatever they want at their next OS
release; apps vendors have no choice but to port to the new release or
die.  Unix vendors can't do that; where standards that affect the
interface between applications and the OS are concerned, they have to
achieve consensus on how that interface can be extended without breaking
existing applications.  While that may slow the adoption of some features
(many of which are arguably more cosmetic than functional anyway), it also
benefits the entire development process by encouraging a focus on
disciplined engineering and standards compliance rather than tricks. 
Remember the TSR nightmares on DOS, and the .DLL hell on Windows (the
latter allegedly mitigated but not eliminated by just adding more
complications)?

[Linux takes a middle road on how much it's willing to break to change, on
the theory that source code is available anyway, so all the apps ought to
be able to keep up, and for the most part, one doesn't really need to be
running the latest version anyway, patches typically being available for
older versions too. That works well enough for some purposes (web servers
and the like), but it remains to be seen how well it works for high-volume
stuff like desktops.  I tend to think that the real world needs both
commercial and open source apps, and the commercial apps darn near require
binary-level backwards compatibility; even with the open source apps, it's
nice to be able to mostly decouple OS updates from apps updates, with just
some routine testing to spot the misbehaving apps.]

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

On Sun, 28 Dec 2003 13:31:18 -0000,
Richard.L.Hamilton@mindwarp.smart.net (Richard L. Hamilton) wrote:

>In article <e5kprvg380asilftdn2dpustmml3hkpi43@4ax.com>,
>	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
>> On Wed, 19 Nov 2003 18:08:12 +0100, "pidbel"
>> <pidbel@pidbel.pidbel.sdf> wrote:
>> 
>>>
>>>"Josh McKee" <jtmckee@rm-bogus-ac.net> wrote in message
>>>news:mr7nrv8lng5tpgtdgu0immbmog8f07a4d1@4ax.com...
>>>> On Tue, 18 Nov 2003 17:33:51 -0800, "Michael Vilain
>>>> <vilain@spamcop.net>" wrote:
>>>>
>>>> And you wonder why Windows is the dominant platform. It may be
>>>> technically inferior to UNIX. But it appears to accommodate the user
>>>> in the areas that are most important to them.
>>>>
>>>> Josh
>>>
>>>Indeed, if you forget the time staring at the blue screen and the daily
>>>security patches you have to install , Windows is accommodating the user
>>>very well ....
>> 
>> Yet Windows still commands a huge share of the market. Why is that?
>> It's because Windows is giving the user what they want. It should be
>> obvious that ease of use is what the users want. Reliability and
>> security are taking a back seat to user friendliness. You can present
>> your case about the strengths of UNIX and the weaknesses of Windows
>> all day long. But in the end the user is going to want the user
>> friendliness of Windows over the reliability of UNIX.
>>
>
>It's fine to be all worried about what the itty bitty poor baby users
>want, but they're like little children; giving them what they want is
>_wrong_ if it could hurt them or burn down the house.

What you fail to understand is that I am not advocating that you
ignore the current 8 character limit and give users a longer user ID.
What I am proposing is increasing the 8 character limit to something
more reasonable. What is reasonable? I don't know but 8 characters
appears to be insufficient in many cases.

>One of those strengths of Unix that you seem to disdain in favor of
>friendliness is that updates to the operating system almost never break
>applications.

You are mistakenly equating my explanation of what users want as my
wants.

>Raising the userid limit above 8 characters is one of those
>things that could potentially break lots of existing applications.
>(if you want to know why it could do that, read the cuserid(3c) man page
>and strain your brain a little...)

So we shouldn't work to improve the system? That's fine...but don't be
surprised when people prefer Windows of UNIX. Windows is giving the
end users what *they* want. That's one of the reasons that it's been
so successful.

>[...]
>> They probably forget their user ID because it's something cryptic
>> designed to fit within an 8 character limitation of the system. It
>> requires the user to adapt to the system and not the other way around.
>> 
>>>So what is the case about the 8 character limitations , as some users
>>>allready can forget his login of one character :-))))
>> 
>> But if it were their name how likely do you think that they'd forget
>> it? Probably unlikely.
>
>There used to be a suggestion a long time ago in one of the admin manuals
>to take their initials and enough letters from the beginning of their last
>name to make eight letters; if that wasn't unique, change the last
>character to a digit.
>
>A procedure like that produces something that's should be easy enough to
>remember. A procedure is also better than letting users pick their own
>names, because the procedure can deal consistently with avoiding conflicts;
>no need to take a lot of crap when the name someone wants is already taken.

Where did I say that users should be allowed to arbitrarily choose
their own user IDs? Isn't using their first initial and full last name
a procedure for creating user IDs? Isn't using their full first and
last name a procedure for creating user IDs?

>Lots of operating systems other than Unix have short user names.  At least
>until the time it picked up a bunch of Unix compatibility features, I think
>at least one of the major IBM mainframe OS's had a seven character limit.
>I recall one other OS (long gone now, but that's not why) that had a six
>character limit.

That's a foolish way of thinking. Just because there is worse doesn't
mean that you shouldn't try to improve.

Josh

On Sun, 28 Dec 2003 14:33:23 -0000,
Richard.L.Hamilton@mindwarp.smart.net (Richard L. Hamilton) wrote:

>In article <7nkprv4g4iqm593grkhvq84rven8i9fap5@4ax.com>,
>	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
>> On Wed, 19 Nov 2003 23:23:00 +0000, Chris Newport
>> <me@see-my-sig.invalid> wrote:
>> 
>>>On Wednesday 19 November 2003 4:54 pm in comp.sys.sun.admin Josh McKee
>>>wrote:
>>>
>>> 
>>>>>> Why are you calling them lusers? Shouldn't the system accommodate the
>>>>>> user instead of the user accommodating the system?
>>>>>
>>>>>In this case, the users must accomodate the system, since it's a hard
>>>>>limit.  If you think it's cost effective to spend time modifying Solaris
>>>>>and all the parts affected by this to accomedate usernames, passwords
>>>>>and groups > 8 characters, then by all means have fun.
>>>> 
>>>> The problem I have is Chris' portrayal of the user as a loser because
>>>> they wish to use something more descriptive. I fail to see why that is
>>>> an unreasonable expectation and warrants the SA calling their
>>>> customers losers. Computers are supposed to work for us, not the other
>>>> way around.
>>>
>>>Oh purleeeese !
>>>
>>>Are you lysdexic ?.
>>>Can you tell the difference between luser and loser ?
>> 
>> No.
>> 
>>>The two terms are not the same, although they often overlap.
>> 
>> Perhaps you'd be willing to explain why loser and luser are not, for
>> all intents and purposes, the same? According to dictionary.com they
>> appear to be synonymous.
>
>Losers are (barring a miracle) losers forever, whereas some lusers may
>earn the right to get the "l" out.
>
>It may not do to tell people to their face that they're probably idiots,
>but as a practical matter, if your job is to maintain something that
>people use, you need to presume that they're going to _act_ like idiots at
>least until such time as specific ones demonstrate otherwise.

I have a problem with an SA thinking that someone is a luser because
they want to use their full name. I don't think such a request is
unreasonable. And since they're most likely not aware of the
limitation why would you consider them an idiot? It's probably not
their job to know...just like it's not your job to know the law or
accounting (I assume that you're neither a lawyer of an accountant) in
detail. Calling someone an idiot because they don't know details about
areas that they're not responsible for just shows what an arrogant ass
you (figuratively speaking) are.

>If you don't think that's true, just spend 5 minutes watching "Jerry
>Springer" or similar shows.  Think of how many pervs and nutcases there
>must be out there to let themselves be made fools of for no more than a
>free airplane ticket, limo ride, a night in a hotel, and a couple of
>meals.  Now think how many more fools there are that watch such garbage to
>feel better about themselves 'cause at least they're not _that_ bad.
>Individually, people will fall on a range from clueless to very clueful
>indeed.  Collectively, although they'll occasionally show a surprising
>ability to detect BS, they'll tend to be dumber according to the inverse
>of the square of the number of people in question, or at least on some
>curve that approaches but never quite reaches the lowest level among
>them.  The very survival of dictators also demonstrates that; they'd never
>stand a chance if even 20% of their people stood against them, and the
>amount of death and misery involved in overthowing them (given that once
>you start, you've got to see it through) would be less than that of a few
>years of their continuing reign.
>
>[...]
>>>The simple solution is to educate the luser and impart the clue
>>>that his email address need not be the same as his login name.
>> 
>> Nothing wrong with that...except that you're classifying them as
>> losers instead of users. I disagree with that. Especially when the
>> computer should accommodate the user and not the user having to
>> accommodate the system. Do you consider that unreasonable?
>
>As compared to breaking backwards compatibility for a bunch of existing
>applications, yes.  Being able to run any number of commercial (and
>in-house) apps that actually do something useful is more important than
>getting the name you want.

Is UNIX so pathetic that this cannot be changed? No, it's not. UNIX
has evolved considerably over the years. Sun took it from 32 bit to 64
bit without much problem. I'm certain that they could, over time, work
to change the 8 character limit. I don't think that this could happen
over night...such changes never do. But if you don't start, you'll
never end.

My question is: Why is everyone so against increasing this limit?

Josh

>It's just imaginable that changes could be made such that user names could
>be allowed to be longer as long as they remained unique within the first
>eight characters for the sake of existing applications incapable of
>handling longer than eight character user names; that could perhaps be
>done without breaking backwards compatibility.  There are some reasons
>to believe that things are slowly heading in that direction.  But AFAIK
>they aren't there yet, so erring on the safe side is only responsible.
>
>There are always constraints, and trade-offs.  People don't take the
>time to repeatedly say "you can't have that" just to be mean (they
>may do it once or twice to be mean, but it gets old too quickly to
>keep that up for long).  They say it because the harm outweighs the
>benefits, at least until such time as a consensus is found on how to
>achieve a reasonable approximation of what's so often requested while
>still honoring functional considerations like backwards compatibility.
>
>Microsoft, being monolithic, can break whatever they want at their next OS
>release; apps vendors have no choice but to port to the new release or
>die.  Unix vendors can't do that; where standards that affect the
>interface between applications and the OS are concerned, they have to
>achieve consensus on how that interface can be extended without breaking
>existing applications.  While that may slow the adoption of some features
>(many of which are arguably more cosmetic than functional anyway), it also
>benefits the entire development process by encouraging a focus on
>disciplined engineering and standards compliance rather than tricks. 
>Remember the TSR nightmares on DOS, and the .DLL hell on Windows (the
>latter allegedly mitigated but not eliminated by just adding more
>complications)?
>
>[Linux takes a middle road on how much it's willing to break to change, on
>the theory that source code is available anyway, so all the apps ought to
>be able to keep up, and for the most part, one doesn't really need to be
>running the latest version anyway, patches typically being available for
>older versions too. That works well enough for some purposes (web servers
>and the like), but it remains to be seen how well it works for high-volume
>stuff like desktops.  I tend to think that the real world needs both
>commercial and open source apps, and the commercial apps darn near require
>binary-level backwards compatibility; even with the open source apps, it's
>nice to be able to mostly decouple OS updates from apps updates, with just
>some routine testing to spot the misbehaving apps.]

In article <0vrtuv8jjhb5ci080ne1an287mjnm6k5cp@4ax.com>, jtmckee@rm-
bogus-ac.net says...
> What you fail to understand is that I am not advocating that you
> ignore the current 8 character limit and give users a longer user ID.
> What I am proposing is increasing the 8 character limit to something
> more reasonable. What is reasonable? I don't know but 8 characters
> appears to be insufficient in many cases.

No matter what length a username is, that length will not be 
satisfactory for some reason.  Eight letters. is too short, and the 
usernames are too cryptic?  Fine, let's make the length twenty letters.  
Whoops, now users cannot remember their usernames!  Damn!  Okay, let's 
make the length twelve letters.  But that's too short for people with 
really long names!  (Round and round we go, where it stops nobody 
knows.)

> So we shouldn't work to improve the system? That's fine...but don't be
> surprised when people prefer Windows of UNIX.

People who whine about petty shit such as how long a username can be 
should just forget all about any *nix system and stick with Windows.  
Those people are just going to 1001 more trivial, shallow so-called 
problems to moan about on a *nix system.

There are far more important things to worry about than the fact that 
your username has to be "jmckee" instead of "josh_mckee".  I mean, 
c'mon.  Next we'll hear moaning about how *nix uses a forward slash in 
directory paths, but Windows uses a backward slash.  How awful!  
Somebody call Ashcroft and beg him to make it go away!

> Windows is giving the end users what *they* want. That's one of the
> reasons that it's been so successful.

It's a question of what users -need-, not what they want.  They "want" 
the entire world handed to them on a silver platter.  They seldom know 
what they really need.  I think Windows is successful for two main 
reasons: (a) Nothing else is mainstream because Microsoft has pretty 
much monopolized the PC OS market, and (b) It has a very low learning 
curve, which is a polite way of saying it appeals to morons.

> Where did I say that users should be allowed to arbitrarily choose
> their own user IDs? Isn't using their first initial and full last name
> a procedure for creating user IDs? Isn't using their full first and
> last name a procedure for creating user IDs?

Oh, but I thought you wanted to give users what they WANT.  They usually 
don't really like the usernames that system admins dream up for them.

But I have a solution that will lay this whole issue to rest right here 
and now.  It will allow usernames of ANY length on a *nix system, yet 
still comply with the eight character length restriction.  That's right, 
boys and girls, ol' Greg has the answer that will make EVERYBODY happy!  
Well, everybody except the system admins...  But anyway, here's what you 
do:  We know that *nix focuses on the first eight characters of a 
username, right?  You can punch in a ten-character username and *nix 
will ignore the last two characters.  (That's really where the eight-
character limit first comes into play.)  This means that you could write 
a policy for the system admins that says they are to assign usernames of 
up to, say, 15 characters but that the first 8 characters must not match 
any other username.  Now the users will get nice, long usernames and the 
system won't get confused!  YES!!  Happy, happy, joy, joy!  (You might 
want to move and change your phone number before the system admins come 
to your house and burn it down.)

On Sun, 28 Dec 2003 10:41:40 -0500, grog <greg@or.y> wrote:

>In article <0vrtuv8jjhb5ci080ne1an287mjnm6k5cp@4ax.com>, jtmckee@rm-
>bogus-ac.net says...
>> What you fail to understand is that I am not advocating that you
>> ignore the current 8 character limit and give users a longer user ID.
>> What I am proposing is increasing the 8 character limit to something
>> more reasonable. What is reasonable? I don't know but 8 characters
>> appears to be insufficient in many cases.
>
>No matter what length a username is, that length will not be 
>satisfactory for some reason.

Perhaps...but don't you think that a user ID length of 32 would likely
be more satisfactory than 8? I'm not saying go crazy and build the
system to accommodate 4096 character user ID's. But it is obvious that
8 characters is to short in many circumstances.

>Eight letters. is too short, and the usernames are too cryptic?  Fine, let's 
>make the length twenty letters.  Whoops, now users cannot remember their
>usernames!

Again...it's unlikely that the user will forget their user name if it
is their name without any truncation. It's easier to remember first
initial full last name than it is first initial seven characters of
their last name.

>Damn!  Okay, let's make the length twelve letters.  But that's too short for people
>with really long names!  (Round and round we go, where it stops nobody 
>knows.)

No...you come up with a standard such as first initial and last name.
It would be foolish, as you're suggesting, to limit user IDs to a
certain number of characters.

>> So we shouldn't work to improve the system? That's fine...but don't be
>> surprised when people prefer Windows of UNIX.
>
>People who whine about petty shit such as how long a username can be 
>should just forget all about any *nix system and stick with Windows.

It seems that they are.
  
>Those people are just going to 1001 more trivial, shallow so-called 
>problems to moan about on a *nix system.

The system should accommodate their needs instead of them having to
accommodate the system. Funny...I thought computers were to help us,
not the other way around.

>There are far more important things to worry about than the fact that 
>your username has to be "jmckee" instead of "josh_mckee".

Apparently not to the users. And if the system cannot accommodate
their needs then maybe the system should be changed to do so. And if
it cannot they'll go somewhere else.

I've said this before but I'll say it again. I've encountered this
problem many times when trying to sell people on UNIX. They see this
limitation, along with a number of other "non-important" (to SAs)
limitations, and opt to go with Windows.

>I mean, c'mon.  Next we'll hear moaning about how *nix uses a forward slash in 
>directory paths, but Windows uses a backward slash.  How awful!  
>Somebody call Ashcroft and beg him to make it go away!
>
>> Windows is giving the end users what *they* want. That's one of the
>> reasons that it's been so successful.
>
>It's a question of what users -need-, not what they want.

It's also a question of what they want. Computers should accommodate
users wants as well as their needs.

>They "want" the entire world handed to them on a silver platter.  They seldom
>know what they really need.

There's that arrogance showing through.

>I think Windows is successful for two main reasons: (a) Nothing else is mainstream
>because Microsoft has pretty much monopolized the PC OS market,

Which they did by giving people what they wanted. They listened to the
user sand made a fortune by giving them what they wanted.

>and (b) It has a very low learning curve, which is a polite way of saying it appeals
>to morons.

I would say just the opposite. Intelligent people chose the easier of
two ways.

>> Where did I say that users should be allowed to arbitrarily choose
>> their own user IDs? Isn't using their first initial and full last name
>> a procedure for creating user IDs? Isn't using their full first and
>> last name a procedure for creating user IDs?
>
>Oh, but I thought you wanted to give users what they WANT.  They usually 
>don't really like the usernames that system admins dream up for them.

I think that most of them are fine with first initial and last name.
What they don't want is first initial and seven characters of their
ten character last name.

>But I have a solution that will lay this whole issue to rest right here 
>and now.  It will allow usernames of ANY length on a *nix system, yet 
>still comply with the eight character length restriction.  That's right, 
>boys and girls, ol' Greg has the answer that will make EVERYBODY happy!  
>Well, everybody except the system admins...  But anyway, here's what you 
>do:  We know that *nix focuses on the first eight characters of a 
>username, right?  You can punch in a ten-character username and *nix 
>will ignore the last two characters.  (That's really where the eight-
>character limit first comes into play.)  This means that you could write 
>a policy for the system admins that says they are to assign usernames of 
>up to, say, 15 characters but that the first 8 characters must not match 
>any other username.  Now the users will get nice, long usernames and the 
>system won't get confused!  YES!!  Happy, happy, joy, joy!  (You might 
>want to move and change your phone number before the system admins come 
>to your house and burn it down.)

Not sure how this would play with directory services.

Josh

In article <p8vtuvsldk8uj3120a6ti4hmmrv6aav927@4ax.com>, jtmckee@rm-
bogus-ac.net says...
> >
> >It's a question of what users -need-, not what they want.
> 
> It's also a question of what they want. Computers should accommodate
> users wants as well as their needs.
> 
> >They "want" the entire world handed to them on a silver platter.  They seldom
> >know what they really need.
> 
> There's that arrogance showing through.

Just speaking from experience.  I've dealt with users constantly for 
years.  They always want pretty much everything (and top quality too!  
and I want it right now!) but they hardly ever know what they actually 
need.

> >But I have a solution that will lay this whole issue to rest right here 
> >and now.  It will allow usernames of ANY length on a *nix system, yet 
> >still comply with the eight character length restriction.  That's right, 
> >boys and girls, ol' Greg has the answer that will make EVERYBODY happy!  
> >Well, everybody except the system admins...  But anyway, here's what you 
> >do:  We know that *nix focuses on the first eight characters of a 
> >username, right?  You can punch in a ten-character username and *nix 
> >will ignore the last two characters.  (That's really where the eight-
> >character limit first comes into play.)  This means that you could write 
> >a policy for the system admins that says they are to assign usernames of 
> >up to, say, 15 characters but that the first 8 characters must not match 
> >any other username.  Now the users will get nice, long usernames and the 
> >system won't get confused!  YES!!  Happy, happy, joy, joy!  (You might 
> >want to move and change your phone number before the system admins come 
> >to your house and burn it down.)
> 
> Not sure how this would play with directory services.

What?  You mean *nix isn't the only software that doesn't readily 
support long usernames?  I never would have guessed!  (Hint: Heavy 
sarcasm there.)

Actually, the problem would involve a lot more than just directory 
services, but those would be issues for the system admins to resolve.  
Might want to keep them away from heavy, sharp, or explosive objects 
though, because they might get uppity.

In article <oqjprv48vb5avg72fafuaac8tsfvpree1n@4ax.com>,
	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
> On Wed, 19 Nov 2003 17:26:01 +0000, Chris Newport
> <me@see-my-sig.invalid> wrote:
> 
>>On Wednesday 19 November 2003 5:04 pm in comp.sys.sun.admin Josh McKee
>>wrote:
>>
>>> On Wed, 19 Nov 2003 01:31:44 +0000, Chris Newport
>>> <me@see-my-sig.invalid> wrote:
>>> 
>>>>On Tuesday 18 November 2003 11:19 pm in comp.sys.sun.admin Josh McKee
>>>>wrote:
>>>>
>>>>> On Tue, 18 Nov 2003 22:55:09 +0000, Chris Newport
>>>>> <me@see-my-sig.invalid> wrote:
>>>>> 
>>>>>>On Tuesday 18 November 2003 9:20 pm in comp.sys.sun.admin Lukas Ruf
>>>>>>wrote:
>>>>>>
>>>>>>> Dear all,
>>>>>>> 
>>>>>>> does anybody know what is the maximal length of usernames on Solairs
>>>>>>> 6,7,8 and above?
>>>>>>> 
>>>>>>> Once, I was told that they are limited to 8 characters.  Is this
>>>>>>> true?
>>>>>>
>>>>>>Thus is not a Solaris issue, it is a Unix standard.
>>>>>>Only the first 8 characters of a username, group, or password
>>>>>>are significant. Everything else should be ignored.
>>>>>>
>>>>>>Be warned that janesmith is the same as janesmit
>>>>>>
>>>>>>Using more than 8 characters is unsupported and often untested.
>>>>>>Breaking the rule can tickle all manner of obscure bugs.
>>>>>>For example, in HPUX 10.20 a long group name caused backups
>>>>>>to mysteriously and silently fail.
>>>>>>
>>>>>>If your lusers want a long email name use a mail alias.
>>>>> 
>>>>> Why are you calling them lusers? Shouldn't the system accommodate the
>>>>> user instead of the user accommodating the system?
>>>>
>>>>Patience, young grasshopper.
>>>>Enlightenment will come with experience.
>>> 
>>> I fail to see how this answered the question. Do you have an answer?
>>
>>The clue has been imparted - 8 characters is the maximum.
> 
> I am well aware of this.
> 
>>This is a hard limit, resistance is futile.
> 
> Resistance? What resistance are you referring to?
> 
>>The enlightened sysadmin does not mess with the standards.
> 
> Who said anything about messing with standards?
> 
>>On that path lies danger.
>>
>>Do you wish to absorb the clue as a step on your path to 
>>enlightenment or are you waiting for someone to impart it
>>with a cluebat after you screw something up ?.
> 
> Who said anything about messing with the standard? I am merely
> inquiring into your classification of users who want longer than 8
> characters in their user ID as losers. I don't think that it is
> unreasonable for a user to want to use what works best for them and
> not what works best for the system. Computers should work for the
> user, not the other way around. If the standard is limiting perhaps
> the standard should be altered to accommodate the users requirements.
> 

Altering the standard would cause major breakage.

The basic limitation is that L_cuserid in /usr/include/stdio.h has been 9
(8 + null byte terminator) for ages, and lots of programs have arrays like

char name[L_cuserid];

in them.  Altering L_cuserid in the include file will not change those
programs, which would still expect no more than 8+1 back from cuserid().
But altering cuserid() to give back more would break those programs,
without warning and in various ways, some of which would lead to data
corruption.

If names were allowed to be longer but still required to be unique within
8 characters, a new function could be added (I propose cluserid(), just
to be annoying :-) that retrieved the long version of the name (and which
new programs could migrate to, keeping in mind that eventually the
uniqueness requirement might be dropped, so they should presently enforce
it when new names are being created but not require it otherwise); and
until such time if any as the unique with 8 characters requirement was
dropped, existing programs could continue to use cuserid().


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

In article <vuu6ptlqqgkl75@corp.supernews.com>, 
Richard.L.Hamilton@mindwarp.smart.net says...
> In article <oqjprv48vb5avg72fafuaac8tsfvpree1n@4ax.com>,
> 	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
[chop] 

That's a troll.  It really has no interest in how long usernames are or 
anything else.  My advice would be to make it dance for you for a while, 
just for laughs, and then plonk it.

In article <jtmckee-090C55.17423524112003@netnews.attbi.com>,
	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
> In article <bptapg$3l8$2@reader2.panix.com>,
>  Jeremiah DeWitt Weiner <jdw@panix.com> wrote:
> 
>> Josh McKee <jtmckee@rm-bogus-ac.net> wrote:
>> > The problem I have is Chris' portrayal of the user as a loser because
>> > they wish to use something more descriptive. 
>> 
>> 	No one commented on this, but I think this is the philosophical
>> crux of the matter.  The username is not supposed to be descriptive,
> 
> Why not? Especially if that's what the users want.
> 
>> it's supposed to be a unique identifier.  If you want descriptive, that's what
>> the GECOS field is for.  
> 
> Users don't log in with their GECOS field.
> 
>> "As a general rule, I am adamantly opposed to using full names as
>> e-mail addresses, since they are not in any sense unique.  For example,
>> the Unix software-development community has two Andy Tannenbaums,
>> at least two well-known Peter Deutsches, and at one time Bell Labs
>> had two Stephen R. Bournes with offices along the same hallway."
>> 	-- Eric Allman
> 
> Eric should learn to listen to the user and not force his opinion on 
> them. The user is the customer. We shouldn't lose sight of that. Phone 
> books listen numbers by full names. Why should e-mail addresses be any 
> different? If there's a discrepency that's easily solved by appending a 
> numerical count to someones name.

You can just as easily take 8 characters from their name and modify
the last to a digit as needed to make it unique.  That's not any harder
to remember, either.

Anyone who really _believes_ the customer is always right has never gone
home after a long day of being polite to idiots and wondered why they
didn't own a punching bag.

One office I was in had one in the machine room (where users did not
enter); it only took a few days to break it.

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

On Sun, 28 Dec 2003 18:04:45 -0000,
Richard.L.Hamilton@mindwarp.smart.net (Richard L. Hamilton) wrote:

>In article <oqjprv48vb5avg72fafuaac8tsfvpree1n@4ax.com>,
>	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
>> On Wed, 19 Nov 2003 17:26:01 +0000, Chris Newport
>> <me@see-my-sig.invalid> wrote:
>> 
>>>On Wednesday 19 November 2003 5:04 pm in comp.sys.sun.admin Josh McKee
>>>wrote:
>>>
>>>> On Wed, 19 Nov 2003 01:31:44 +0000, Chris Newport
>>>> <me@see-my-sig.invalid> wrote:
>>>> 
>>>>>On Tuesday 18 November 2003 11:19 pm in comp.sys.sun.admin Josh McKee
>>>>>wrote:
>>>>>
>>>>>> On Tue, 18 Nov 2003 22:55:09 +0000, Chris Newport
>>>>>> <me@see-my-sig.invalid> wrote:
>>>>>> 
>>>>>>>On Tuesday 18 November 2003 9:20 pm in comp.sys.sun.admin Lukas Ruf
>>>>>>>wrote:
>>>>>>>
>>>>>>>> Dear all,
>>>>>>>> 
>>>>>>>> does anybody know what is the maximal length of usernames on Solairs
>>>>>>>> 6,7,8 and above?
>>>>>>>> 
>>>>>>>> Once, I was told that they are limited to 8 characters.  Is this
>>>>>>>> true?
>>>>>>>
>>>>>>>Thus is not a Solaris issue, it is a Unix standard.
>>>>>>>Only the first 8 characters of a username, group, or password
>>>>>>>are significant. Everything else should be ignored.
>>>>>>>
>>>>>>>Be warned that janesmith is the same as janesmit
>>>>>>>
>>>>>>>Using more than 8 characters is unsupported and often untested.
>>>>>>>Breaking the rule can tickle all manner of obscure bugs.
>>>>>>>For example, in HPUX 10.20 a long group name caused backups
>>>>>>>to mysteriously and silently fail.
>>>>>>>
>>>>>>>If your lusers want a long email name use a mail alias.
>>>>>> 
>>>>>> Why are you calling them lusers? Shouldn't the system accommodate the
>>>>>> user instead of the user accommodating the system?
>>>>>
>>>>>Patience, young grasshopper.
>>>>>Enlightenment will come with experience.
>>>> 
>>>> I fail to see how this answered the question. Do you have an answer?
>>>
>>>The clue has been imparted - 8 characters is the maximum.
>> 
>> I am well aware of this.
>> 
>>>This is a hard limit, resistance is futile.
>> 
>> Resistance? What resistance are you referring to?
>> 
>>>The enlightened sysadmin does not mess with the standards.
>> 
>> Who said anything about messing with standards?
>> 
>>>On that path lies danger.
>>>
>>>Do you wish to absorb the clue as a step on your path to 
>>>enlightenment or are you waiting for someone to impart it
>>>with a cluebat after you screw something up ?.
>> 
>> Who said anything about messing with the standard? I am merely
>> inquiring into your classification of users who want longer than 8
>> characters in their user ID as losers. I don't think that it is
>> unreasonable for a user to want to use what works best for them and
>> not what works best for the system. Computers should work for the
>> user, not the other way around. If the standard is limiting perhaps
>> the standard should be altered to accommodate the users requirements.
>> 
>
>Altering the standard would cause major breakage.

Perhaps if it wasn't done correctly. Many limitations in UNIX have
increased over the years. I fail to see why this one couldn't be
increased as well.

Josh

>The basic limitation is that L_cuserid in /usr/include/stdio.h has been 9
>(8 + null byte terminator) for ages, and lots of programs have arrays like
>
>char name[L_cuserid];
>
>in them.  Altering L_cuserid in the include file will not change those
>programs, which would still expect no more than 8+1 back from cuserid().
>But altering cuserid() to give back more would break those programs,
>without warning and in various ways, some of which would lead to data
>corruption.
>
>If names were allowed to be longer but still required to be unique within
>8 characters, a new function could be added (I propose cluserid(), just
>to be annoying :-) that retrieved the long version of the name (and which
>new programs could migrate to, keeping in mind that eventually the
>uniqueness requirement might be dropped, so they should presently enforce
>it when new names are being created but not require it otherwise); and
>until such time if any as the unique with 8 characters requirement was
>dropped, existing programs could continue to use cuserid().

On Sun, 28 Dec 2003 13:24:48 -0500, grog <greg@or.y> wrote:

>In article <vuu6ptlqqgkl75@corp.supernews.com>, 
>Richard.L.Hamilton@mindwarp.smart.net says...
>> In article <oqjprv48vb5avg72fafuaac8tsfvpree1n@4ax.com>,
>> 	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
>[chop] 
>
>That's a troll.

I often hear that from people who cannot form any tangible response to
what I have to say. What's even more amazing is their need to
interrupt the intelligent conversations others are having with me so
that they can proclaim to that person that I am a troll. Oh well,
children will never learn.

>It really has no interest in how long usernames are or 
>anything else.  My advice would be to make it dance for you for a while, 
>just for laughs, and then plonk it.

Then wouldn't you be a troll?

Josh

In article <bpfhl5$kiq$1@news.ox.ac.uk>,
	Graham Lee <graham.lee@wadham.oxford.ac.invalid.uk> writes:
> Josh McKee wrote:
> 
>> On Tue, 18 Nov 2003 22:55:09 +0000, Chris Newport
>> <me@see-my-sig.invalid> wrote:
>> 
>>>
>>>If your lusers want a long email name use a mail alias.
>> 
>> Why are you calling them lusers? Shouldn't the system accommodate the
>> user instead of the user accommodating the system?
>> 
> 
> <offtopic>An old BOFH-ism, which I think goes back to a hacked report
> program (similar to finger, I expect) on an ITS system.  It went from
> saying "x users logged in" when you pressed C-z, to saying "x losers logged
> in".  Some users, apparently, didn't like being called losers, so the
> compromise "lusers" was reached.  It stuck.
> 
> ISTR the help command on ITS was duly renamed "luser", too :)</offtopic>
> 
> I've just gone in and hacked the passwd stuff on a passing Solaris 8
> machine.  If you have a luser with a stupid name (such as 'ihavelongname'),
> then they *can* log in and *can* change their password.  I did so and the
> updated passwd entry doesn't get truncated.  What you don't seem able to
> do, however, is to log in as 'ihavelon'; the 8-character truncated version
> of 'ihavelongname'.
> 
> *HOWEVER*, that's not to say you should do it.  Who knows how big the buffer
> is in programs such as login, passwd, etc; usernames are designed to be 8
> characters long and one shouldn't rock the boat.  Also, who knows what kind
> of problems it could lead to should you require your Solaris systems to
> interoperate with anyone else's UNIX.

It boils down to a buffer overrun on anything that depends on the existing
value of L_cuserid in stdio.h.

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

In article <slrnbrmdef.v0e.ruf_block_spam@komsys-pc-ruf.ethz.ch>,
	Lukas Ruf <ruf_block_spam@rawip.org> writes:
> Greg Andrews wrote:
>>  me@see-my-sig.invalid writes:
>> >On Tuesday 18 November 2003 9:20 pm in comp.sys.sun.admin Lukas Ruf
>> >wrote:
>> >
>> >> does anybody know what is the maximal length of usernames on
>> >> Solairs 6,7,8 and above?
>> >>
>>
>>  Solaris can be said to support long usernames because nothing in
>>  the base OS breaks on them.  The "bugs" which people refer to are
>>  almost always things like information in the output of ps not being
>>  in nice columns, as with short names.  I call these annoyances
>>  rather than bugs.
> 
> Thanks for the clarification!
> 
> I do not care about cosmetic annoyances when it comes to system
> reliability and stability.
> 
>>
>>  Some other programs (that you download from the Internet rather
>>  than come with Solaris) may break, but I haven't run into any yet.
>>
> 
> As long as the base system -- i.e. more specifically: the kernel,
> mountd, automountd, nis+ etc. -- is running stable, I am fine with it!
> 
>>  It doesn't have to be a problem, as long as you can live with the
>>  annoyances that long usernames produce.
>>
> 
> I can, definitely! ;-)
> 
>>  (note I've only commented on usernames, not on passwords or other
>>  things)
>>
> 
> what's the issue there?
> 
> Thanks for any further enlightenment!
> 

See the man page for the function cuserid(), and the constant L_cuserid
in stdio.h.  Any program compiled with L_cuserid as 9 (8 + null byte
terminator) could be broken in very strange ways, which might not
be obvious yet could result in eventual data corruption, by anything longer
than 8 character account names.

Even if nothing in Solaris depends on that anymore (and I don't know for a
fact whether that's true), plenty of other commercial and in-house
software might.

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

On Sun, 28 Dec 2003 19:08:54 -0000,
Richard.L.Hamilton@mindwarp.smart.net (Richard L. Hamilton) wrote:

>In article <jtmckee-090C55.17423524112003@netnews.attbi.com>,
>	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
>> In article <bptapg$3l8$2@reader2.panix.com>,
>>  Jeremiah DeWitt Weiner <jdw@panix.com> wrote:
>> 
>>> Josh McKee <jtmckee@rm-bogus-ac.net> wrote:
>>> > The problem I have is Chris' portrayal of the user as a loser because
>>> > they wish to use something more descriptive. 
>>> 
>>> 	No one commented on this, but I think this is the philosophical
>>> crux of the matter.  The username is not supposed to be descriptive,
>> 
>> Why not? Especially if that's what the users want.
>> 
>>> it's supposed to be a unique identifier.  If you want descriptive, that's what
>>> the GECOS field is for.  
>> 
>> Users don't log in with their GECOS field.
>> 
>>> "As a general rule, I am adamantly opposed to using full names as
>>> e-mail addresses, since they are not in any sense unique.  For example,
>>> the Unix software-development community has two Andy Tannenbaums,
>>> at least two well-known Peter Deutsches, and at one time Bell Labs
>>> had two Stephen R. Bournes with offices along the same hallway."
>>> 	-- Eric Allman
>> 
>> Eric should learn to listen to the user and not force his opinion on 
>> them. The user is the customer. We shouldn't lose sight of that. Phone 
>> books listen numbers by full names. Why should e-mail addresses be any 
>> different? If there's a discrepency that's easily solved by appending a 
>> numerical count to someones name.
>
>You can just as easily take 8 characters from their name and modify
>the last to a digit as needed to make it unique.  That's not any harder
>to remember, either.

I'm not talking about uniqueness. I'm talking about why someone with
the name of "Richard Hamilton" can't have a user ID such as
"rhamilton" instead of "rhamilto". I notice that your e-mail address
is "richard.l.hamilton". Why aren't you using "rhamilto" instead?

>Anyone who really _believes_ the customer is always right has never gone
>home after a long day of being polite to idiots and wondered why they
>didn't own a punching bag.

I never said that the customer was always right. I said that the user
is the customer and that we shouldn't lose sight of that. If they want
longer user ID's then reasonable steps should be taken to accommodate
their needs. Microsoft saw this and has profited handsomely as a
result.

Josh

>One office I was in had one in the machine room (where users did not
>enter); it only took a few days to break it.

In article <0vrtuv8jjhb5ci080ne1an287mjnm6k5cp@4ax.com>,
	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
> On Sun, 28 Dec 2003 13:31:18 -0000,
> Richard.L.Hamilton@mindwarp.smart.net (Richard L. Hamilton) wrote:
[...]
>>Lots of operating systems other than Unix have short user names.  At least
>>until the time it picked up a bunch of Unix compatibility features, I think
>>at least one of the major IBM mainframe OS's had a seven character limit.
>>I recall one other OS (long gone now, but that's not why) that had a six
>>character limit.
> 
> That's a foolish way of thinking. Just because there is worse doesn't
> mean that you shouldn't try to improve.

Where did I say that _if_ a safe and standards-consensus-based way of
expanding the userid limit were developed, it shouldn't be used? I just
don't want one Unix vendor rushing into a solution that's not portable to
other Unix implementations, nor do I want people taking advantage of the
fact that even now it may not break the OS without thinking about how it
might break other applications.  Nor, on a list of prioritized future
features under consideration, do I consider longer account names as
important as quite a few other possibilities, which isn't to say that it
might not be nice to have, just that there's rightly plenty ahead of it on
the list.

Anyway, if one generates account names by a rule, one should reasonably
choose a rule that satisfies the strictest constraints of any of the
systems on which one might be creating accounts.  That's why one such rule
I've seen uses a seven character limit, even though that's shorter than
what Windows (or Unix) allow, because they still have some systems with
other OS's that expect the seven character limit, and it's more handy to
have one rule for everything than different rules for each different OS
they use.

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

In article <p8vtuvsldk8uj3120a6ti4hmmrv6aav927@4ax.com>,
	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
> On Sun, 28 Dec 2003 10:41:40 -0500, grog <greg@or.y> wrote:
> 
>>In article <0vrtuv8jjhb5ci080ne1an287mjnm6k5cp@4ax.com>, jtmckee@rm-
>>bogus-ac.net says...
>>> What you fail to understand is that I am not advocating that you
>>> ignore the current 8 character limit and give users a longer user ID.
>>> What I am proposing is increasing the 8 character limit to something
>>> more reasonable. What is reasonable? I don't know but 8 characters
>>> appears to be insufficient in many cases.
>>
>>No matter what length a username is, that length will not be 
>>satisfactory for some reason.
> 
> Perhaps...but don't you think that a user ID length of 32 would likely
> be more satisfactory than 8? I'm not saying go crazy and build the
> system to accommodate 4096 character user ID's. But it is obvious that
> 8 characters is to short in many circumstances.
> 
>>Eight letters. is too short, and the usernames are too cryptic?  Fine, let's 
>>make the length twenty letters.  Whoops, now users cannot remember their
>>usernames!
> 
> Again...it's unlikely that the user will forget their user name if it
> is their name without any truncation. It's easier to remember first
> initial full last name than it is first initial seven characters of
> their last name.

Negligibly so.  All you really have to remember is that it's what you want,
only shorter.

>>Damn!  Okay, let's make the length twelve letters.  But that's too short for people
>>with really long names!  (Round and round we go, where it stops nobody 
>>knows.)
> 
> No...you come up with a standard such as first initial and last name.
> It would be foolish, as you're suggesting, to limit user IDs to a
> certain number of characters.
> 
>>> So we shouldn't work to improve the system? That's fine...but don't be
>>> surprised when people prefer Windows of UNIX.
>>
>>People who whine about petty shit such as how long a username can be 
>>should just forget all about any *nix system and stick with Windows.
> 
> It seems that they are.
>   
>>Those people are just going to 1001 more trivial, shallow so-called 
>>problems to moan about on a *nix system.
> 
> The system should accommodate their needs instead of them having to
> accommodate the system. Funny...I thought computers were to help us,
> not the other way around.
> 
>>There are far more important things to worry about than the fact that 
>>your username has to be "jmckee" instead of "josh_mckee".
> 
> Apparently not to the users. And if the system cannot accommodate
> their needs then maybe the system should be changed to do so. And if
> it cannot they'll go somewhere else.

What part of "that's not a _need_" don't you understand.  It's a _want_.
Yeah, well I want my system to fetch me a beer, and take the cap off too
(I like the expensive bottled stuff, that generally doesn't have screw-off
caps).  Why don't I go look for an OS that'll do that for me?  Sheesh!

> I've said this before but I'll say it again. I've encountered this
> problem many times when trying to sell people on UNIX. They see this
> limitation, along with a number of other "non-important" (to SAs)
> limitations, and opt to go with Windows.
> 
>>I mean, c'mon.  Next we'll hear moaning about how *nix uses a forward slash in 
>>directory paths, but Windows uses a backward slash.  How awful!  
>>Somebody call Ashcroft and beg him to make it go away!
>>
>>> Windows is giving the end users what *they* want. That's one of the
>>> reasons that it's been so successful.
>>
>>It's a question of what users -need-, not what they want.
> 
> It's also a question of what they want. Computers should accommodate
> users wants as well as their needs.
> 
>>They "want" the entire world handed to them on a silver platter.  They seldom
>>know what they really need.
> 
> There's that arrogance showing through.

No, that's a fact.  They don't think about this stuff for a living, they
just use it.  Now, if this were totally trivial and could be done quickly
with no adverse consequences, I suppose some more of us would agree with
you.  But it isn't, and because doing it properly means changing the
standard (which is a matter of consensus and cannot be dictated by any one
company, government, power, or deity), it can't happen quickly.  There's
already an RFE in for it on Solaris, AFAIK, it's just not high enough on
the list to have gotten very far, esp. since a proper solution is a
standards based one and not just something Sun goes off and does on their
own.

>>I think Windows is successful for two main reasons: (a) Nothing else is mainstream
>>because Microsoft has pretty much monopolized the PC OS market,
> 
> Which they did by giving people what they wanted. They listened to the
> user sand made a fortune by giving them what they wanted.
> 
>>and (b) It has a very low learning curve, which is a polite way of saying it appeals
>>to morons.
> 
> I would say just the opposite. Intelligent people chose the easier of
> two ways.

Intelligent people invest an effort in learning how to use their tools
proportionate to the amount of flexibility they anticipate needing in
their uses of the tool.  That means that if all you do is data entry, then
for all I care you can use Windows; but if your work involves doing
something almost every day that's a bit different than what you've done
before, then I think Windows may not be your best choice.  Windows is
memorize the menus; Unix is learn the tools.  The latter is more flexible,
obviously.

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

On Sun, 28 Dec 2003 19:56:48 -0000,
Richard.L.Hamilton@mindwarp.smart.net (Richard L. Hamilton) wrote:

>In article <0vrtuv8jjhb5ci080ne1an287mjnm6k5cp@4ax.com>,
>	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
>> On Sun, 28 Dec 2003 13:31:18 -0000,
>> Richard.L.Hamilton@mindwarp.smart.net (Richard L. Hamilton) wrote:
>[...]
>>>Lots of operating systems other than Unix have short user names.  At least
>>>until the time it picked up a bunch of Unix compatibility features, I think
>>>at least one of the major IBM mainframe OS's had a seven character limit.
>>>I recall one other OS (long gone now, but that's not why) that had a six
>>>character limit.
>> 
>> That's a foolish way of thinking. Just because there is worse doesn't
>> mean that you shouldn't try to improve.
>
>Where did I say that _if_ a safe and standards-consensus-based way of
>expanding the userid limit were developed, it shouldn't be used?

It's implied because you have been arguing against my proposal to
increase the length of the user ID in UNIX.

>I just don't want one Unix vendor rushing into a solution that's not portable to
>other Unix implementations, nor do I want people taking advantage of the
>fact that even now it may not break the OS without thinking about how it
>might break other applications.  Nor, on a list of prioritized future
>features under consideration, do I consider longer account names as
>important as quite a few other possibilities, which isn't to say that it
>might not be nice to have, just that there's rightly plenty ahead of it on
>the list.
>
>Anyway, if one generates account names by a rule, one should reasonably
>choose a rule that satisfies the strictest constraints of any of the
>systems on which one might be creating accounts.  That's why one such rule
>I've seen uses a seven character limit, even though that's shorter than
>what Windows (or Unix) allow, because they still have some systems with
>other OS's that expect the seven character limit, and it's more handy to
>have one rule for everything than different rules for each different OS
>they use.

That's easier said than done. Often times administrators of one
platform may not be familiar with all the limitations of other
platforms.

Josh

On Sun, 28 Dec 2003, Richard L. Hamilton wrote:

> It boils down to a buffer overrun on anything that depends on the existing
> value of L_cuserid in stdio.h.

You are correct.  If all programs could be changed to use
the run-time limit (sysconf (_SC_LOGIN_NAME_MAX)), rather
than a compile time constant list L_cuserid, increasingthe
length of a username wouldn't be so painful.

But that is VERY unlikely to happen, so for better or worse,
UNIX is essentially "stuck" with an 8 character username
limit.

-- 
Rich Teer, SCNA, SCSA                               .  *   * . * .* .
                                                     .   *   .   .*
President,                                          * .  . /\ ( .  . *
Rite Online Inc.                                     . .  / .\   . * .
                                                    .*.  / *  \  . .
                                                      . /*   o \     .
Voice: +1 (250) 979-1638                            *   '''||'''   .
URL: http://www.rite-online.net                     ******************

On Sun, 28 Dec 2003, Josh McKee wrote:

> their needs. Microsoft saw this and has profited handsomely as a
> result.

I very much doubt that M$'s success on the desktop is
due to the fact that Windoze allows long login names!
Windoze had a monopoly long before Windoze even had the
concept of usernames...

-- 
Rich Teer, SCNA, SCSA                               .  *   * . * .* .
                                                     .   *   .   .*
President,                                          * .  . /\ ( .  . *
Rite Online Inc.                                     . .  / .\   . * .
                                                    .*.  / *  \  . .
                                                      . /*   o \     .
Voice: +1 (250) 979-1638                            *   '''||'''   .
URL: http://www.rite-online.net                     ******************

In article <9heuuv07uc2rtl1v6l6a924l5b935umm2p@4ax.com>,
	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
> On Sun, 28 Dec 2003 19:56:48 -0000,
> Richard.L.Hamilton@mindwarp.smart.net (Richard L. Hamilton) wrote:
> 
>>In article <0vrtuv8jjhb5ci080ne1an287mjnm6k5cp@4ax.com>,
>>	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
>>> On Sun, 28 Dec 2003 13:31:18 -0000,
>>> Richard.L.Hamilton@mindwarp.smart.net (Richard L. Hamilton) wrote:
>>[...]
>>>>Lots of operating systems other than Unix have short user names.  At least
>>>>until the time it picked up a bunch of Unix compatibility features, I think
>>>>at least one of the major IBM mainframe OS's had a seven character limit.
>>>>I recall one other OS (long gone now, but that's not why) that had a six
>>>>character limit.
>>> 
>>> That's a foolish way of thinking. Just because there is worse doesn't
>>> mean that you shouldn't try to improve.
>>
>>Where did I say that _if_ a safe and standards-consensus-based way of
>>expanding the userid limit were developed, it shouldn't be used?
> 
> It's implied because you have been arguing against my proposal to
> increase the length of the user ID in UNIX.

Maybe someone is, I haven't seen it.  I'm arguing that (a) I don't think
it's as important as you do (not irrelevant, just not as important), and
(b) that user wants are like children's wants: if the resources are there,
and it can be done without harm, then when there's a chance to do it
right, fine.  Otherwise, be sensitive the first time you say no, and firm
thereafter.  Never explain unless they accept the "no" first, because
persuasion by reasoning with someone whose frame of reference on the
issues is so different accomplishes nothing.

>>I just don't want one Unix vendor rushing into a solution that's not portable to
>>other Unix implementations, nor do I want people taking advantage of the
>>fact that even now it may not break the OS without thinking about how it
>>might break other applications.  Nor, on a list of prioritized future
>>features under consideration, do I consider longer account names as
>>important as quite a few other possibilities, which isn't to say that it
>>might not be nice to have, just that there's rightly plenty ahead of it on
>>the list.
>>
>>Anyway, if one generates account names by a rule, one should reasonably
>>choose a rule that satisfies the strictest constraints of any of the
>>systems on which one might be creating accounts.  That's why one such rule
>>I've seen uses a seven character limit, even though that's shorter than
>>what Windows (or Unix) allow, because they still have some systems with
>>other OS's that expect the seven character limit, and it's more handy to
>>have one rule for everything than different rules for each different OS
>>they use.
> 
> That's easier said than done. Often times administrators of one
> platform may not be familiar with all the limitations of other
> platforms.

Organizational IT policies need to be hashed out at a level slightly (IMO
not _too_ much, mind you; this CIO is God's viceroy crap gets old fast)
above that of single-platform administrators.

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

On Sun, 28 Dec 2003 22:36:04 GMT, Rich Teer <rich.teer@rite-group.com>
wrote:

>On Sun, 28 Dec 2003, Josh McKee wrote:
>
>> their needs. Microsoft saw this and has profited handsomely as a
>> result.
>
>I very much doubt that M$'s success on the desktop is
>due to the fact that Windoze allows long login names!

I didn't say that this was the *only* reason. I said that it was one
of the reasons. Why is this such a difficult concept for people to
understand. It isn't anyone *one* thing...it's a combination of *many*
little [inconsequential in some peoples opinion] things. Microsoft
listen to the user about what they wanted and they provided it. Not
just one thing...many things. Understand? Didn't think so.

Josh

>Windoze had a monopoly long before Windoze even had the
>concept of usernames...

On Sun, 28 Dec 2003 22:32:36 GMT, Rich Teer <rich.teer@rite-group.com>
wrote:

>On Sun, 28 Dec 2003, Richard L. Hamilton wrote:
>
>> It boils down to a buffer overrun on anything that depends on the existing
>> value of L_cuserid in stdio.h.
>
>You are correct.  If all programs could be changed to use
>the run-time limit (sysconf (_SC_LOGIN_NAME_MAX)), rather
>than a compile time constant list L_cuserid, increasingthe
>length of a username wouldn't be so painful.
>
>But that is VERY unlikely to happen, so for better or worse,
>UNIX is essentially "stuck" with an 8 character username
>limit.

Of course it's unlikely to happen...because people like you don't
think it's beneficial to increase it. That's the point: You're not
listening to what the customer wants.

Josh

On Mon, 29 Dec 2003, Josh McKee wrote:

> Of course it's unlikely to happen...because people like you don't
> think it's beneficial to increase it. That's the point: You're not

I've never said it wasn't beneficial.  FWIW, I think it would be
a good idea if the length of login names could be increased--provided
it can be done in a manner that doesn't break backwards compatibility.

To reiterate: I think longer login names are desirable, but not so
desirable that they're worth breaking backwards compatibility for.

Oh, and my car does have a heater.  That's because Chrysler managed
to add a heater to it without changing how I expect to drive it.

-- 
Rich Teer, SCNA, SCSA                               .  *   * . * .* .
                                                     .   *   .   .*
President,                                          * .  . /\ ( .  . *
Rite Online Inc.                                     . .  / .\   . * .
                                                    .*.  / *  \  . .
                                                      . /*   o \     .
Voice: +1 (250) 979-1638                            *   '''||'''   .
URL: http://www.rite-online.net                     ******************

In article <dubuuvsa3i6vm7e27eih0cbr05dpk2o879@4ax.com>,
	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
> On Sun, 28 Dec 2003 19:08:54 -0000,
> Richard.L.Hamilton@mindwarp.smart.net (Richard L. Hamilton) wrote:
> 
>>In article <jtmckee-090C55.17423524112003@netnews.attbi.com>,
>>	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
>>> In article <bptapg$3l8$2@reader2.panix.com>,
>>>  Jeremiah DeWitt Weiner <jdw@panix.com> wrote:
>>> 
>>>> Josh McKee <jtmckee@rm-bogus-ac.net> wrote:
>>>> > The problem I have is Chris' portrayal of the user as a loser because
>>>> > they wish to use something more descriptive. 
>>>> 
>>>> 	No one commented on this, but I think this is the philosophical
>>>> crux of the matter.  The username is not supposed to be descriptive,
>>> 
>>> Why not? Especially if that's what the users want.
>>> 
>>>> it's supposed to be a unique identifier.  If you want descriptive, that's what
>>>> the GECOS field is for.  
>>> 
>>> Users don't log in with their GECOS field.
>>> 
>>>> "As a general rule, I am adamantly opposed to using full names as
>>>> e-mail addresses, since they are not in any sense unique.  For example,
>>>> the Unix software-development community has two Andy Tannenbaums,
>>>> at least two well-known Peter Deutsches, and at one time Bell Labs
>>>> had two Stephen R. Bournes with offices along the same hallway."
>>>> 	-- Eric Allman
>>> 
>>> Eric should learn to listen to the user and not force his opinion on 
>>> them. The user is the customer. We shouldn't lose sight of that. Phone 
>>> books listen numbers by full names. Why should e-mail addresses be any 
>>> different? If there's a discrepency that's easily solved by appending a 
>>> numerical count to someones name.
>>
>>You can just as easily take 8 characters from their name and modify
>>the last to a digit as needed to make it unique.  That's not any harder
>>to remember, either.
> 
> I'm not talking about uniqueness. I'm talking about why someone with
> the name of "Richard Hamilton" can't have a user ID such as
> "rhamilton" instead of "rhamilto". I notice that your e-mail address
> is "richard.l.hamilton". Why aren't you using "rhamilto" instead?

I did that 'cause I saw one of the Sun guys doing it and was curious how
they did it, not that I care whether it looks better, or is easier for
anyone to remember, or type, or whatever.  My account name is
rlhamil; email to either one is equivalent, including for all follow-on
processing such as spam filtering, automatic filing into separate topical
mailboxes, etc.  Outgoing mail gets mapped to Richard.L.Hamilton (case is
apparently irrelevant); incoming mail gets mapped to rlhamil.  If there
were a reason to, and it were a matter for my discretion (it isn't at
work), I could set it up for someone else to do that too, now that I've
done it once.  Big deal.  That's not an account name, per se.  I wouldn't
want it as an account name; too much to type right and fast.  I may have
to type my login name hundreds of times in a day, I don't need to deal
with even a few percent more errors, let alone even a few more characters
to type.  I suppose for someone who only typed it once a day, the blazing
obviousness of their full name (in most cases where there was no suffix
for uniqueness) might outweigh the shortness, but after a couple of weeks,
it's hard for me to imagine that that someone would be anything like me.

Besides, decades before that, my account was just my initials (on one
system; on another it was some arbitrary thing made up, etc.;ISTR some
rather ugly account name on Burroughs MCS, although I never had any
problems with it at the time).  If one day all the systems I have accounts
on supported a 32 character limit and there was a shift to an account name
formula that took advantage of that, that'd be ok too, although all other
things being equal I'd prefer _not_ to have something longer than what I
have now. It's not like we're issued account names at birth; we _are_
issued Social Security numbers (in the US) as soon as our parents want to
claim us as a tax deduction, but using a Social Security number as an
account name (even though it's unique) would be a Really Bad Idea. Our
mere names are not guaranteed to be unique; I know mine isn't. So 32 may
be a little bit better than 8, but we're not at the point where a single
rule will formulate an account name that will work for the rest of our
lives.  So I personally don't get much benefit from having a longer
limit.  A little benefit?  Yeah, maybe if I cared whether or not other
people could remember it (I don't, since if I don't already know them I
don't care a whit whether I ever hear from them).  I don't terribly care
what my account name is, as long as it's not anything rude or silly;
except on systems that are my personal property, it's not like it was
something I had much choice about, even if they were systems I
administered.

>>Anyone who really _believes_ the customer is always right has never gone
>>home after a long day of being polite to idiots and wondered why they
>>didn't own a punching bag.
> 
> I never said that the customer was always right. I said that the user
> is the customer and that we shouldn't lose sight of that. If they want
> longer user ID's then reasonable steps should be taken to accommodate
> their needs. Microsoft saw this and has profited handsomely as a
> result.

You mixed wants and needs again; and that aside, I'm not satisfied with
the argument that MS got rich from being responsive.  I think it had a lot
to do also with being in the right place at the right time, and engaging
in some very sharp dealing and heavy-handed tactics that were ok for a
small company but that they kept on with well past that point, and also
with the hardware vs OS history.  Windows 3.1, pathetic as it was, ran ok
on a 286;Unix with a decent GUI AFAIK wasn't really attempted on Intel
until the 386 (SCO Xenix with an X server, I think that would've been). 
Yes, some Unix flavors ran on more or less commodity systems as early as
the XT (e.g. PC/IX, done for IBM by ISC), but they were slow, and had damn
near no GUI.  And when Unix went from free to government and universities
(pre-AT&T breakup, when they weren't allowed to sell software) to
commercial, it was way too expensive, which didn't change until Windows
was already overwhelmingly dominant.  There are plenty of other factors
too, I'm sure.  There may be some sense in which MS's responsiveness put
them ahead of other companies (certainly their responsiveness to IBM put
them ahead of say, Digital Research!).  But I don't find the mere claim
that their responsiveness to the end user's wants or needs was mainly what
put them at the head of the pack to be persuasive.  Indeed, I'm reasonably
confident that I could come up with hundreds if not thousands of anecdotes
to the contrary (not that that would prove anything either). Maybe there's
a satisfaction study somewhere that compares MS and say Apple (and maybe
Linux, too).  That'd be mildly interesting, _if_ it were statistically
sound, and _if_ it wasn't affected (plus or minus) by the simple fact that
so many more people are familiar with Windows (as a user, at least) than
with anything else.


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

In article <Pine.SOL.4.58.0312281429300.9814@zaphod.rite-group.com>,
	Rich Teer <rich.teer@rite-group.com> writes:
> On Sun, 28 Dec 2003, Richard L. Hamilton wrote:
> 
>> It boils down to a buffer overrun on anything that depends on the existing
>> value of L_cuserid in stdio.h.
> 
> You are correct.  If all programs could be changed to use
> the run-time limit (sysconf (_SC_LOGIN_NAME_MAX)), rather
> than a compile time constant list L_cuserid, increasingthe
> length of a username wouldn't be so painful.
> 
> But that is VERY unlikely to happen, so for better or worse,
> UNIX is essentially "stuck" with an 8 character username
> limit.

Well, I described a way (or two) how it could be done (longer
name unique in 8 characters, or a new long name in addition to
the existing short name, where either would work with new or modified
programs, and only the short name would work with unmodified
programs; ambiguities to be avoided, and among domains subject
to routine re-orgs (like components of a larger organization),
any given longname should only have one shortname and vice versa.
Either way, cuserid() still only honors established L_cuserid,
and a new API hands out the full name, subject to the runtime limit.
Technically possible; somewhat of a pain, but that's manageable too.

utmpx ut_user is 32 in Solaris (for probably over ten years now,
oldest thing I have handy is 2.5.1 and it's almost that) so it's arguably
more than half done.  Even fixing all the rest of the Solaris software
wouldn't be all that hard.  The hard part would be coming up with
a prototype, and getting enough vendors on board to get it through
POSIX and TOG's SUS.  Not so much a technical problem (up to a point)
as a resource and a quasi-political problem.

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

On Mon, 29 Dec 2003 01:21:55 GMT, Rich Teer <rich.teer@rite-group.com>
wrote:

>On Mon, 29 Dec 2003, Josh McKee wrote:
>
>> Of course it's unlikely to happen...because people like you don't
>> think it's beneficial to increase it. That's the point: You're not
>
>I've never said it wasn't beneficial.  FWIW, I think it would be
>a good idea if the length of login names could be increased--provided
>it can be done in a manner that doesn't break backwards compatibility.

Do I have to explicitly state that the method should not break
backwards compatibility? Or do I have to assume that you're an idiot
and spell it out for you? It should be obvious that backwards
compatibility should be a goal of the change *without the need to
explicitly state it*.

>To reiterate: I think longer login names are desirable, but not so
>desirable that they're worth breaking backwards compatibility for.

Show me where i have *EVER* stated that it should be done regardless
of the impact to backwards compatibility. Try reading and
understanding what I've written instead of reading into it what you
want.

>Oh, and my car does have a heater.  That's because Chrysler managed
>to add a heater to it without changing how I expect to drive it.

So what...having a heater is a want. It's not needed so why do you
expect to get one? After all...you could wear a coat. And I'm certain
that you would tell Chrysler to take a hike if they told you that you
don't need a heater because you could wear a coat. See how stupid the
"it's a want and not a need" point is? Meeting the customers wants is
sound business advice. I'm amazed that such a simple concept is beyond
you.

Josh

On Mon, 29 Dec 2003 02:10:27 -0000,
Richard.L.Hamilton@mindwarp.smart.net (Richard L. Hamilton) wrote:

>In article <dubuuvsa3i6vm7e27eih0cbr05dpk2o879@4ax.com>,
>	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
>> On Sun, 28 Dec 2003 19:08:54 -0000,
>> Richard.L.Hamilton@mindwarp.smart.net (Richard L. Hamilton) wrote:
>> 
>>>In article <jtmckee-090C55.17423524112003@netnews.attbi.com>,
>>>	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
>>>> In article <bptapg$3l8$2@reader2.panix.com>,
>>>>  Jeremiah DeWitt Weiner <jdw@panix.com> wrote:
>>>> 
>>>>> Josh McKee <jtmckee@rm-bogus-ac.net> wrote:
>>>>> > The problem I have is Chris' portrayal of the user as a loser because
>>>>> > they wish to use something more descriptive. 
>>>>> 
>>>>> 	No one commented on this, but I think this is the philosophical
>>>>> crux of the matter.  The username is not supposed to be descriptive,
>>>> 
>>>> Why not? Especially if that's what the users want.
>>>> 
>>>>> it's supposed to be a unique identifier.  If you want descriptive, that's what
>>>>> the GECOS field is for.  
>>>> 
>>>> Users don't log in with their GECOS field.
>>>> 
>>>>> "As a general rule, I am adamantly opposed to using full names as
>>>>> e-mail addresses, since they are not in any sense unique.  For example,
>>>>> the Unix software-development community has two Andy Tannenbaums,
>>>>> at least two well-known Peter Deutsches, and at one time Bell Labs
>>>>> had two Stephen R. Bournes with offices along the same hallway."
>>>>> 	-- Eric Allman
>>>> 
>>>> Eric should learn to listen to the user and not force his opinion on 
>>>> them. The user is the customer. We shouldn't lose sight of that. Phone 
>>>> books listen numbers by full names. Why should e-mail addresses be any 
>>>> different? If there's a discrepency that's easily solved by appending a 
>>>> numerical count to someones name.
>>>
>>>You can just as easily take 8 characters from their name and modify
>>>the last to a digit as needed to make it unique.  That's not any harder
>>>to remember, either.
>> 
>> I'm not talking about uniqueness. I'm talking about why someone with
>> the name of "Richard Hamilton" can't have a user ID such as
>> "rhamilton" instead of "rhamilto". I notice that your e-mail address
>> is "richard.l.hamilton". Why aren't you using "rhamilto" instead?
>
>I did that 'cause I saw one of the Sun guys doing it and was curious how
>they did it, not that I care whether it looks better, or is easier for
>anyone to remember, or type, or whatever.

Then go back to using your user ID instead. There's less overhead on
the system.

>My account name is rlhamil; email to either one is equivalent, including for all follow-on
>processing such as spam filtering, automatic filing into separate topical
>mailboxes, etc.  Outgoing mail gets mapped to Richard.L.Hamilton (case is
>apparently irrelevant); incoming mail gets mapped to rlhamil.  If there
>were a reason to, and it were a matter for my discretion (it isn't at
>work), I could set it up for someone else to do that too, now that I've
>done it once.  Big deal.  That's not an account name, per se.  I wouldn't
>want it as an account name; too much to type right and fast.  I may have
>to type my login name hundreds of times in a day, I don't need to deal
>with even a few percent more errors, let alone even a few more characters
>to type.  I suppose for someone who only typed it once a day, the blazing
>obviousness of their full name (in most cases where there was no suffix
>for uniqueness) might outweigh the shortness, but after a couple of weeks,
>it's hard for me to imagine that that someone would be anything like me.
>
>Besides, decades before that, my account was just my initials (on one
>system; on another it was some arbitrary thing made up, etc.;ISTR some
>rather ugly account name on Burroughs MCS, although I never had any
>problems with it at the time).  If one day all the systems I have accounts
>on supported a 32 character limit and there was a shift to an account name
>formula that took advantage of that, that'd be ok too, although all other
>things being equal I'd prefer _not_ to have something longer than what I
>have now. It's not like we're issued account names at birth; we _are_
>issued Social Security numbers (in the US) as soon as our parents want to
>claim us as a tax deduction, but using a Social Security number as an
>account name (even though it's unique) would be a Really Bad Idea. Our
>mere names are not guaranteed to be unique; I know mine isn't. So 32 may
>be a little bit better than 8, but we're not at the point where a single
>rule will formulate an account name that will work for the rest of our
>lives.  So I personally don't get much benefit from having a longer
>limit.  A little benefit?  Yeah, maybe if I cared whether or not other
>people could remember it (I don't, since if I don't already know them I
>don't care a whit whether I ever hear from them).  I don't terribly care
>what my account name is, as long as it's not anything rude or silly;
>except on systems that are my personal property, it's not like it was
>something I had much choice about, even if they were systems I
>administered.
>
>>>Anyone who really _believes_ the customer is always right has never gone
>>>home after a long day of being polite to idiots and wondered why they
>>>didn't own a punching bag.
>> 
>> I never said that the customer was always right. I said that the user
>> is the customer and that we shouldn't lose sight of that. If they want
>> longer user ID's then reasonable steps should be taken to accommodate
>> their needs. Microsoft saw this and has profited handsomely as a
>> result.
>
>You mixed wants and needs again;

Not at all. I am specifically referencing wants because the customers
wants are important. It's not your position, or mine, to determine
what is most important to the user. It's theirs. And if we tell them
enough times that their wants are unimportant then they'll go to
someone who will provide for their wants. Period.

Josh

 and that aside, I'm not satisfied with
>the argument that MS got rich from being responsive.  I think it had a lot
>to do also with being in the right place at the right time, and engaging
>in some very sharp dealing and heavy-handed tactics that were ok for a
>small company but that they kept on with well past that point, and also
>with the hardware vs OS history.  Windows 3.1, pathetic as it was, ran ok
>on a 286;Unix with a decent GUI AFAIK wasn't really attempted on Intel
>until the 386 (SCO Xenix with an X server, I think that would've been). 
>Yes, some Unix flavors ran on more or less commodity systems as early as
>the XT (e.g. PC/IX, done for IBM by ISC), but they were slow, and had damn
>near no GUI.  And when Unix went from free to government and universities
>(pre-AT&T breakup, when they weren't allowed to sell software) to
>commercial, it was way too expensive, which didn't change until Windows
>was already overwhelmingly dominant.  There are plenty of other factors
>too, I'm sure.  There may be some sense in which MS's responsiveness put
>them ahead of other companies (certainly their responsiveness to IBM put
>them ahead of say, Digital Research!).  But I don't find the mere claim
>that their responsiveness to the end user's wants or needs was mainly what
>put them at the head of the pack to be persuasive.  Indeed, I'm reasonably
>confident that I could come up with hundreds if not thousands of anecdotes
>to the contrary (not that that would prove anything either). Maybe there's
>a satisfaction study somewhere that compares MS and say Apple (and maybe
>Linux, too).  That'd be mildly interesting, _if_ it were statistically
>sound, and _if_ it wasn't affected (plus or minus) by the simple fact that
>so many more people are familiar with Windows (as a user, at least) than
>with anything else.

In article <l56vuv4sscr2604gk0js52pgn5g97e5qjo@4ax.com>,
	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
> On Mon, 29 Dec 2003 02:10:27 -0000,
> Richard.L.Hamilton@mindwarp.smart.net (Richard L. Hamilton) wrote:
[...]
>>I did that 'cause I saw one of the Sun guys doing it and was curious how
>>they did it, not that I care whether it looks better, or is easier for
>>anyone to remember, or type, or whatever.
> 
> Then go back to using your user ID instead. There's less overhead on
> the system.

Hey, my system, WTF overhead?  But no problem, it's changed back. (still
works either way incoming so's not to break anything I haven't bothered to
check for, won't rewrite it on the way out anymore; that's two different
things anyway.  That's to make the point.  I may change it back; then
again, I may not; if I stop using two different but equivalent email
addresses on Usenet, presumably I'll stop getting two copies of some of
the spams my filters don't catch.)


> Not at all. I am specifically referencing wants because the customers
> wants are important. It's not your position, or mine, to determine
> what is most important to the user. It's theirs. And if we tell them
> enough times that their wants are unimportant then they'll go to
> someone who will provide for their wants. Period.

As an abstract concept, that's sweet.  In the real world, there's plenty
of times I'd be willing to tell them why I thought (to be polite) they
might wish to readjust their priorities, and, if only as a last resort,
ignore them to the extent need to keep things working smoothly.  I think
that like a doctor, "first, do no harm" at least sometimes takes
precedence over "let the customer determine what's important to them".
I think there are occasions where there's room for (and even a need for)
responsible discretion to be excercised in such matters.

Users (customers) do three things:

* they ask for stuff

* they break things

* one way or the other, internal or external, without any customers,
  we wouldn't be getting paid

There's definitely _some_ connection between how we deal with the first
two and the third, but I maintain that there's definitely some room for
discretion too (and that a high degree of responsiveness to the first one
isn't the majority or probably even the plurality reason for MS's
financial success).  I also maintain that there are plenty of times when
what the user _thinks_ they want isn't what they'd want if you walked them
through the ramifications, and there are some times when it's best not to
unhesitatingly give them what they want even if you haven't had (or given
sufficient prior similar examples providing experience and precedent on
which to draw on, taken) the opportunity to indulge them with the
opportunity to hear such an explanation.  Like anything else in the real
world, there are bound to be some exceptions, but there's a certain amount
of wiggle room most of the time.  It shouldn't be used lightly or
arbitrarily, but it shouldn't be ignored under the illusion of operating
according to some higher market-based principle, even though such
principles do have some meaning.  In the case of your item in question,
their request should be given nonzero consideration and service.  But not
much compared to most other requests or forseeable anticipations of
requests.  And numbers of requests alone, although they provide guidance,
don't dictate priorities absolutely.  A request with an explanation,
compared to the explanations known or reasonably presumed for other
requests, counts for more than an unjustified request, other factors being
equal.  And there may be plenty of reasons to allocate resources to
activities not directly related to user requests that they don't know
about, and unless your satisfaction rate (a nontrivial metric but not a
dictatorial absolute) drops low enough to set off alarms, don't _need_ to
know about.

I don't care in the slightest if you, or anyone that isn't more or less
directly responsible for _my_ salary, thinks that's arrogant, or
unrealistic, or anything else.

I also don't particularly care if most people think me anywhere from
undiplomatically blunt to self-destructively rude and presumptious in
regarding unknown persons as no better than they've proven themselves to
be, even though I might not tell them that to their face.  In my
experience, while I don't want to be utterly unapproachable, I don't want
to be _too_ approachable either.

I'm getting too approachable here...I think I'm going to get some sleep.

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

In article <5pstuvgha0978c0dsu729dprtj1r2uos95@4ax.com>,
	Josh McKee <jtmckee@rm-bogus-ac.net> writes:
> On Sun, 28 Dec 2003 14:33:23 -0000,
> Richard.L.Hamilton@mindwarp.smart.net (Richard L. Hamilton) wrote:
[...]
>>Losers are (barring a miracle) losers forever, whereas some lusers may
>>earn the right to get the "l" out.
>>
>>It may not do to tell people to their face that they're probably idiots,
>>but as a practical matter, if your job is to maintain something that
>>people use, you need to presume that they're going to _act_ like idiots at
>>least until such time as specific ones demonstrate otherwise.
> 
> I have a problem with an SA thinking that someone is a luser because
> they want to use their full name. I don't think such a request is
> unreasonable. And since they're most likely not aware of the
> limitation why would you consider them an idiot? It's probably not
> their job to know...just like it's not your job to know the law or
> accounting (I assume that you're neither a lawyer of an accountant) in
> detail. Calling someone an idiot because they don't know details about
> areas that they're not responsible for just shows what an arrogant ass
> you (figuratively speaking) are.

I'll take the arrogant part, but not the other word.  I _am_ arrogant. I'm
smarter than 3/4 of the people I work with, and I know it.  Which is not
to say they're dummies, most of them aren't (I have my doubts about a
couple...).  Also, don't assume I've never worked in at least one of the
specialties that the users work in; I have (although I admit I'd get my
tail kicked in a number of the others, at least for a few months; I'm not
prepared to concede there's anything physically possible that I couldn't
make a pretty big dent in learning in a year or so if I really wanted
to).  As for the rest, the request in the abstract isn't unreasonable, but
if I've treated someone with the respect they deserve (which I actually do
sometimes...), they'll know that when I say "maybe one day it'll work to
do that, but no for now" that's a fair enough characterization that
they'll just drop it.  And unless all the systems a user already might
encounter in their organization allowed longer names and a policy of name
selection that took advantage of that was already in place, I'd consider
the notion of choosing a system on the basis of that particular feature
alone to be incredibly ignorant unless the systems were otherwise for all
intents and purposes identical.

[...]

> Is UNIX so pathetic that this cannot be changed? No, it's not. UNIX
> has evolved considerably over the years. Sun took it from 32 bit to 64
> bit without much problem. I'm certain that they could, over time, work
> to change the 8 character limit. I don't think that this could happen
> over night...such changes never do. But if you don't start, you'll
> never end.

Actually, it sneaked up on 64-bit rather than going all at once; there
was large file support (> 2GB) in what, Solaris 2.5.1?  Besides, 64-bit
was an opportunity for a do-over on the binary ABI, which is why limits
on the number of stdio file pointers and the number of fds select() can
handle are higher in 64-bit.  One could ask why they didn't increase the
account limit for 64-bit as well, although since 32-bit programs still
exist on 64-bit systems, and there apparently wasn't any interest in
the standards for a longer account name at that time, it might well
be a pointless question.

> My question is: Why is everyone so against increasing this limit?

Well, _if_ increasing the limit can be done without causing cuserid(3c) to
return longer names, and without any other breakage of
non-OS-vendor-supplied software that was previously considered "well
behaved", and if it's blessed by incorporation in the relevant standards,
then I'm ok with it, although not at a priority that would put it ahead of
a number of other IMO more important issues.  Is it possible within those
constraints?  If the name still had to be unique within the first eight
characters, I'd certainly think so.  So one day, ok, but until then,
there's what Alan said about being added to RFE 4109819, or there's
killing time by arguing just for the heck of it.

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

Richard.L.Hamilton@mindwarp.smart.net (Richard L. Hamilton) writes:

>utmpx ut_user is 32 in Solaris (for probably over ten years now,
>oldest thing I have handy is 2.5.1 and it's almost that) so it's arguably
>more than half done.  Even fixing all the rest of the Solaris software
>wouldn't be all that hard.  The hard part would be coming up with
>a prototype, and getting enough vendors on board to get it through
>POSIX and TOG's SUS.  Not so much a technical problem (up to a point)
>as a resource and a quasi-political problem.

Solaris 2.0 already had utmpx but still had utmp; Solaris 7 was the
last release with "/etc/utmp"; Solaris 8 and later only use utmpx.


Usernames of upto 32 characters "mostly work"; but there's some
pesky standards issues (LOGNAME_MAX) that are hard to fix.

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.