Linux-Advocacy Digest #277, Volume #31 Fri, 5 Jan 01 19:13:02 EST
Contents:
Re: linux does NOT suck (oh yes it does) ("Kyle Jacobs")
Re: Why Hatred? (Form@C)
Re: Nobody wants Linux because it destroys hard disks. ("Kyle Jacobs")
Re: Why Hatred? (T. Max Devlin)
Re: Why Hatred? (T. Max Devlin)
Re: Linux, it is great. (The Ghost In The Machine)
----------------------------------------------------------------------------
From: "Kyle Jacobs" <[EMAIL PROTECTED]>
Crossposted-To: alt.linux.sux
Subject: Re: linux does NOT suck (oh yes it does)
Date: Fri, 05 Jan 2001 23:37:33 GMT
You might want to pitch this idea to the LSB.
linuxbase.org
"Ayende Rahien" <[EMAIL PROTECTED]> wrote in message
news:935j2c$69c$[EMAIL PROTECTED]...
>
> "Roberto Alsina" <[EMAIL PROTECTED]> wrote in message
> news:935baf$5ob$[EMAIL PROTECTED]...
> > In article <9356ot$rqr$[EMAIL PROTECTED]>,
> > "Ayende Rahien" <[EMAIL PROTECTED]> wrote:
> > >
> > > "Roberto Alsina" <[EMAIL PROTECTED]> wrote in message
> > > news:932c5b$ju6$[EMAIL PROTECTED]...
> > > > In article <9306l2$5sc$[EMAIL PROTECTED]>,
> > > > "Ayende Rahien" <[EMAIL PROTECTED]> wrote:
> > >
> > > > > > > > Because although it knows the options change, it has no
> > > > > > > > idea of HOW they change. For example, if optionA is now
> > > > > > > > deprecated, and those who set optionA to valueA should now
> > > > > > > > set optionB to valueB, it won't know, and it will (in Gus
> > > > > > > > Grissom's words) screw the pooch.
> > > > > > >
> > > > > > > The program doesn't need to remember the settings, it read
> > > > > > > them from the file.
> > > > > >
> > > > > > And the setting is still the old one.
> > > > >
> > > > > No, because when the program installed itself, it would either
> > > > > replace the file with a new, default one, or update the file
> > > > > format.
> > > >
> > > > So, you have the choice of losing the configuration or hope there's
> > > > a converting tool.
> > >
> > > Of course there would be.
> >
> > Never seen one, really. Why would program writers do it?
>
> Assuming that there would be a common file format, there would be GPL/LGPL
> functions/classes/programs to read/write/convert it.
>
> > > See below.
> > >
> > > > > > > A new version will update the file, and should covert the old
> > > > > > > file format to the new format with more/less/updated options.
> > > > > >
> > > > > > No program ever comes with config file updaters.
> > > > >
> > > > > How hard would it be to write such a thing.
> > > > > Or be backward compatible?
> > > >
> > > > If you just expect them to be backward compatible, you end as
> > > > smb.conf, with a bazillion options that work and are more or less
> > > > the same.
> > >
> > > No, I don't expect it to be backward compatible, actually.
> >
> > Then I don't understand what you wrote.
> >
> > > I expect the program makers to be able to read the file and change
> > > the file format to include the new options & won't include the old
> > > settings which are no longer include in the new version.
> >
> > You are expecting the program maker to write a config file converter.
> > I still see no reason why they will bother. Not to mention that doing
> > so becomes more difficult each version.
>
> No, I'm expecting it to use a GPL/LGPL tool to convert it.
>
> > > > And it can indeed be a terribly difficult thing to do.
> > >
> > > I disagree.
> >
> > You know, I *did* have to write such a thing (krnconvert, check it out
> > in KRN 0.6.9, it didn't quite work) and it was a nightmare. Now, why
> > do you say it will be easy? Is it just a belief?
>
> I'm assuming wide-spread use of this file format, if that would be the
case,
> there would be GPL/LGPL ways to deal with it.
> If that would be the case, then a programmer would only need to fetch
those
> functions & use them.
> Reading from the file itself is a non-trivial task compare to the fscanf
you
> need to read from a flat text file that linux programs use today.
> I'm assuming the existance of ways to deal with it.
>
> > > > > > The format is the same. The options are not.
> > > > >
> > > > > I know, so?
> > > > > It's the program's fault for keeping an old, no-longer-supported
> > > > > format, as its config file.
> > > >
> > > > Who cares? The user is still left with a non-functional setup.
> > >
> > > Whose fault is it?
> >
> > I DON'T CARE! In these issues, knowing whose fault it is and a nickel
> > will get you a small coin.
> >
> > > If I install a new program I fully expect it to be able to read the
> > > old format and covert it to the new one.
> >
> > You haven't installed much software, have you?
>
> Only a couple of thousands of applications.
>
> > > > And this is really the easy part, still. What happens if you have
> > > > interdependent settings? For instance, What happens if a global
> > > > default setting changes the options available for others?
> > > >
> > > > For example:
> > > >
> > > > if A is "a", then B can be "1" or "2".
> > > > if A is "b", then B can be "3" or "4".
> > > > if A is "c", then B must be "5".
> > >
> > > Check the file, it now has dependencies check.
> >
> > What file?
>
> http://www10.ewebcity.com/ayende/lmc.xml
>
> > > I'll improve it sometimes this week.
> >
> > You are shifting the complexity around. You moved the dependency
> > checking from the config tool into the file format. Now the file
> > format is less human readable. Writing a config file from scratch
> > is now almost impossible. Way to go!
>
> I think that a common file format such as the one in the file above is
worth
> the disadvantages you mentioned.
> Someone else could write a better file format, more human readable.
>
> > Complexity doesn't usually vanish in the air, and I see no signs of
> > this particular instance doing that.
>
> Userwise, it does.
> The user can configure all his programs from one tool, and I don't mean
> pico, and he can see all the options that are opened to him.
> Personally, I think it's worth the trouble, you apperantly disagree.
>
>
------------------------------
Subject: Re: Why Hatred?
From: [EMAIL PROTECTED] (Form@C)
Date: Fri, 05 Jan 2001 23:47:39 GMT
T. Max Devlin <[EMAIL PROTECTED]> wrote in
<[EMAIL PROTECTED]>:
<snip>
>
>The reason for asking question 1 (which I didn't ask, BTW) seems to me
>to be to point out that Win32 is such a crappy specification to begin
>with, that even if MS opened its source and published all of its
>documentation, it still would not become a "standard", by which I
>assumed me meant a *de facto standard*, which is a non-standardized
>specification that is widely implemented across the industry. Win32,
>"Windows", if you will, cannot possibly rate, until there is someone
>besides Microsoft producing it.
>
I realise that it wasn't your question originally! I've been enjoying this
thread...
In this case Linux also, by your own argument, cannot become a "de facto
standard" as there is, effectively, a sole producer of the kernel (Linus
has a "benign monopoly"). The fact that the source is open is irrelevent.
Linux is not produced to any standard specification AFAIK. If it is, it
can't be to the same standard as unix, Minix etc. There appear to be too
many differences.
Also, Win32 is not actually a "specification" anyway! It is, however,
"widely implemented across the industry" even if it is technically inferior
to Linux, which is far less commonly implemented at present. Incidentally,
when you mention the "Win32" specification, which one are you referring to,
the desktop or the server? They are two very different products which share
a common UI.
The Win32(desktop) "specification" does not seek to attain the same goals
as Linux, the only area in which they even remotely compete is on server
platforms. The two can, quite legitimately, co-exist because of this
(well, W9x (on desktops) and Linux (on servers) can).
I am sure that there are many cases of the "wrong" specification becoming
an industry "standard". This generally happens because the inferior product
has one or two very redeeming features. e.g. VHS video casettes became
popular not because they gave good performance (they didn't), but because
they were very easy for people to copy! Win32(desktop) would probably
become *an* industry standard because of the user interface facilities as
these have a very high priority for its target market.
--
Mick
Olde Nascom Computers - http://www.mixtel.co.uk
------------------------------
From: "Kyle Jacobs" <[EMAIL PROTECTED]>
Crossposted-To: alt.linux.sux
Subject: Re: Nobody wants Linux because it destroys hard disks.
Date: Fri, 05 Jan 2001 23:51:03 GMT
Now now, some people prefer the simplicity and power of the Outlook News
reader.
He may very well be a UNIX "Systems Engineer" who prefers to do his posting
on a Windows machine.
Or you could be 100% right, and he could just be a whiny little troll who
likes to bash Windows because he thinks that "Big Billy" is the one who
forced him to run Windows all these years.
"Ayende Rahien" <[EMAIL PROTECTED]> wrote in message
news:935ibc$5o9$[EMAIL PROTECTED]...
>
> "Kyle Jacobs" <[EMAIL PROTECTED]> wrote in message
> news:BQp56.151406$[EMAIL PROTECTED]...
> > You have to understand that "UNIX Engineers" like Mr. Kulkis would be
out
> of
> > a job if people believed that Windows 2000 was an improvement over
Windows
> > NT 4.
>
> Actually, Mr. Kulkis probably never even *seen* a Unix.
> All his posting were made from a win98 machine.
>
>
------------------------------
From: T. Max Devlin <[EMAIL PROTECTED]>
Subject: Re: Why Hatred?
Reply-To: [EMAIL PROTECTED]
Date: Fri, 05 Jan 2001 23:57:04 GMT
Said Pete Goodwin in comp.os.linux.advocacy on Fri, 5 Jan 2001 07:51:31
>T. Max Devlin wrote:
>
>> That would depend on how good the extensions were, and how good the
>> ideas were. An API that's constantly being extended to incorporate new
>> ideas would be something of a monstrous nightmare, as an API, if the
>> extensions were not well designed, or the ideas were not particularly
>> good to begin with.
>
>I did say "not always wisely".
You didn't say "unacceptable, substandard, and rather pathetic", though,
which is more accurately the case.
>> From my own experience, abandoning my oath to avoid dichotomies, I can
>> characterize all the developers I've heard or read entirely and
>> completely into two groups:
>>
>> A) Those who think Windows is wonderful, Win32 is wonderful,
>> monopolization is wonderful (or at least inevitable and beneficial), and
>> generally (possibly coincidentally) can't hold up their end of a
>> reasoned argument or otherwise engage in reasonable debate in explaining
>> why, but inevitably resort to arguments from ignorance, second-guessing
>> the market, and generally spouting Microsoft hype in place of factual
>> results.
>>
>> B) Those who agree that Windows APIs as a whole are generally monstrous
>> nightmares.
>
>I think you need more groups, because I certainly don't think WIN32 is
>wonderful, and neither do I believe the Windows API is a monstrous
>nightmare.
So you say, but not many people appear to believe you when you say it.
It may well be your personal opinion, it may even be an honest one
(though that seems unlikely; more probably it is that you are a
passive-aggressive person, and wish to remove any contested ground from
the discussion with false compromise.) But it is certainly difficult to
believe, given the circumstances.
--
T. Max Devlin
*** The best way to convince another is
to state your case moderately and
accurately. - Benjamin Franklin ***
Sign the petition and keep Deja's archive alive!
http://www2.PetitionOnline.com/dejanews/petition.html
------------------------------
From: T. Max Devlin <[EMAIL PROTECTED]>
Subject: Re: Why Hatred?
Reply-To: [EMAIL PROTECTED]
Date: Sat, 06 Jan 2001 00:01:09 GMT
Said Erik Funkenbusch in comp.os.linux.advocacy on Fri, 5 Jan 2001
>"T. Max Devlin" <[EMAIL PROTECTED]> wrote in message
>news:[EMAIL PROTECTED]...
>> Said Erik Funkenbusch in comp.os.linux.advocacy on Thu, 4 Jan 2001
>> >Fine, Linux can't replace windows on the majority of users systems (yet).
>>
>> That's where your wrong. According to your own "any script on Unix can
>> be done on NT" logic, Linux is perfectly capable of entirely and
>> completely replacing Windows on every user's system. It just can't *be*
>> Windows.
>
>What???? Your logic is flawed.
How so? You have stated that any script that can be written on Unix can
be written on NT (despite the falsity of this claim to begin with) under
the "its just software" argument, saying 'perl and bash are available
for Windows'. By the same token, of course, any OS is a complete
replacement for Windows. There is nothing incomplete about Linux.
>> >Perhaps someday it can, but Microsoft isn't known for standing still.
>>
>> No, its known for breaking the law and establishing the largest criminal
>> monopoly the world has ever known, actually.
>
>IBM was in anti-trust trials for 20 years.
And didn't have even a small amount of the success at monopolizing that
Microsoft did, and they've only been in existence for a bit over 20
years. IBM didn't *start out* monopolizing, as MS did.
[...]
>> MS didn't "replace" competitors; it destroyed competition. Microsoft
>> constantly churns its product; inside communications reveal that this is
>> done knowingly to prevent competition on the merits, despite the
>> knowledge that it is a detriment to the consumer.
>
>If it was a detriment to the consumer, the consumer wouldn't buy the
>upgrade.
So it is, so it must be. And you wonder why I bring up Microsoft's
constant efforts to force consumers to upgrade, despite their
reluctance.
[...]
>> >I can say that easily because I also can see
>> >competant competitors that don't let MS get a foothold, such as Intuit and
>> >AOL.
>>
>> Hyuk. If only we were all so naive.
>
>Why don't you try addressing the comment instead. Can you counter this
>statement with any facts?
No, because the statement is not fact; it is supposition. You presume
that because there are some companies which ostensibly compete with
Microsoft, therefore Microsoft cannot be monopolizing. Its a vapid
assumption, not a statement of fact, that "competent competitors that
don't let MS get a foothold". In point of fact, it would seem that
*certain* competitors have not been entirely destroyed by the monopoly
(and more than adequate reasons can be seen in the characteristics of
their market themselves), and you wish to presume that this is the
definition of competence, and by that pretend to "prove" that competent
competitors are not destroyed by anti-competitive monopolies, and
therefore all that are destroyed are simply incompetent. Its silly,
really. No, the continued existence of Intuit and AOL does not provide
any working support for hypothesizing, let alone concluding, that
Microsoft does not act anti-competitively, thereby raising prices above
competitive levels and excluding alternatives.
[...]
>> >The Amiga died not because of competition, but because it's parent company
>> >was incompetant. [...]
>>
>> Enough with the drivel.
>
>You're going to claim that Commodore was competant? Give me a break.
It was popular. I leave all other speculation (outside of my own
opinion as an owner that it was a very good system) to those who prefer
to second-guess the incompetence of others, so that they may try to
convince themselves of their own competence, by proxy.
>> >Windows is in fact a reliable, efficient and effective for me. [...]
>>
>> I noticed the subtle yet glaring ad hoc qualification, there, Erik.
>> What's up with that?
>
>Because it's an opinion about my useage.
Bullshit; its an "escape hatch" to make your opinion unfalsifiable. It
also makes it meaningless and arbitrary, but you would have to have some
intellectual integrity to care about that.
>> >I don't claim inexpensive, but then again, it's awfully expensive to
>retrain lots of
>> >people.
>>
>> Huh, what? Is this the excuse? NT is more expensive than WinDOS
>> because of 'retraining' costs? Retraining who? And what possible
>> relationship would that have causing NT to necessarily cost several
>> times what WinDOS does?
>
>No, Linux is more expensive than NT because of training costs. NT/2000
>costs more because it does more. It's a workstation level OS rather than a
>consumer level one.
Bwah-ha-ha-ha.
>> I think its awfully shameful that people don't recognize NT pricing for
>> what it is; a complete sham. Its a scam; a way to increase prices. NT
>> software doesn't inherently *cost* any more to produce.
>
>No, but NT most certainly did cost a lot more to produce than Win9x did (by
>produce I mean R&D, and maintenance).
Yes, but to even suggest that this is important is to bring up the idea
that a product's price is based on the cost of development. But that
would make the price scalable, so that WinDOS would cost about $2 today,
and NT would already be down into double digits. Oops. Maybe you
should try a different fib, er, I mean argument.
>> Microsoft
>> doesn't include the cost of any support for the product, there are no
>> outrageously exorbitant OEM license fees that MS has to pay, the CDs are
>> not a thousand times greater in quality. The only thing that makes it
>> more reliable is its replacement of DOS with this VMS-like thing to
>> support the Win32 API.
>
>Retail copies of NT/2000 include support in the price. OEM copies do not,
>which is why OEM's pay less (in addition to volume pricing).
Per-processor licensing agreements are not "volume pricing". Microsoft
has been in the habit of referring to them as "volume discounts" since
they were shown to be illegal, but the term you're looking for is
"cliff's-edge pricing", which refers to the actual cost to the OEM of
providing monopoly crapware, without which they would go out of
business. They pay less because if MS tried to charge them any more, it
would break their pre-load lock-in, which is a fundamental support for
Microsoft's illegal monopoly.
>> Microsoft, as I've noted, has been rather busy lately coming up with
>> ways to force consumers to pay more for the product that MS says they
>> want everyone to buy, NT/2K/Whatever, to replace WinDOS. But any
>> profit-seeking company can tell them how to do that; charge less money
>> for it. Less then you are, hell, less than WinDOS. It doesn't actually
>> *cost* any more.
>
>It does cost more. You are forgetting about the cost to develop, which must
>be amortized over the life of the product. MS has spent many billions
>developing Win2k and Whistler. That has to be paid for.
No, it doesn't. Because it already was; the "profits" from DOS alone
were more than enough to pay for them. Its a scam, you putz. The cost
to develop is not related to the price of the product, or else the cost
would go down over time, and it hasn't.
[...]
>> >What you labor under is the assumption that people *WANT* to learn about
>> >computers.
>>
>> Some do, some don't. Some will at any one time. Most will at some
>> times, and won't at others. I am under no assumption that it is to my
>> benefit to second-guess the market.
>
>You certainly had no problem second-gussing the market when making your
>original comment.
No, that was probably double-checking. The difference is potentially
subtle and always abstract, so I'm not sure you you'll be able to
understand it. I'm not at all sure what you're pretending to refer to
as my "original comment", though, so if you'll be more explicit, I'll
try to explain it to you.
>> >The vast majority of people who use computers don't want to know
>> >how or why it works. They just want it to work.
>>
>> And here we get to the problem.
>
>Most peoples computers do work, be it Windows, MacOS, or Linux.
You can say that some of the peoples computers work all of the time, and
all of the peoples computers work some of the time, but you can't say
that all of the peoples computers work all of the time. So designing
computers around the premise that they will usually work, and thus have
no capability of continuing to function adequately if anything at all
should go wrong in the slightest, is, well, only fit for monopolists.
Anyone who might require that their products be able to compete with
alternatives and still provide profits are generally going to recognize
that they sometimes fail, and attempt to make such failures as minimally
problematic for the consumer as possible. This is generally why
obfuscation of software code, for instance, is generally considered a
bad thing, even as developers make jokes about 'job security'.
>> >Each release of a MS OS
>> >get's closer and closer to the ideal of the user not having to know much of
>> >anything in order to install and use it and it's applications. Linux
>> >distros, on the other hand take very small steps (if any at all).
>>
>> That's because the "closer and closer" that "each release" of an MS OS
>> is supposed to be taking has been covered with existing computing
>> technology for decades. The "ideal" is to not have an "ideal"; there's
>> no one right way to run a computer, just the way you want to, and
>> possibly the way someone else does.
>
>No, there isn't one right way, but in order to have consistency, one way
>must be chosen. Consistency is the basis for reducing training and support.
Consistency does not mean everything must work identically; it means
everything must work according to the same fundamental mechanisms. This
is the argument that Unix people have for avoiding krufty "bolt-on"
development the way Microsoft practices it. Microsoft's approach is to
pretend (in addition to pretending that nothing ever fails, as stated
above) that the fundamental mechanisms of how a computer works is
unimportant to successfully using a computer. Its delusional, but it
serves the monopoly well. Also, despicable people like you that trumpet
the value of ignorance in order to promote the monopoly.
>> Linux distros don't *have* the kinds of problems that MS OSes have
>
>Thats not true. Linux has it's own form of DLL hell with dynamic library
>dependancy conflicts.
Library conflicts are only theoretically related to DLL hell. They are
problems which affect related parts of an OS. The cause of library
conflicts is free market development; the correction is knowledge. DLL
hell is caused by design deficiencies in the Windows platform; the
correction is limited to ignoring it (the Microsoft way) or abandoning
Windows, and potentially all monopoly crapware, entirely.
>> regardless of your confusion based
>> on the problems which you report that you did have. I don't care if you
>> have problems; everyone has problems.
>
>You just said Linux doesn't have them.
No, I said it doesn't have the class of problems you were referring to.
I said that the problems you experience weren't Linux problems, but Erik
problems.
>> Those who use MS OSes have more
>> problems, whether its contrary to your personal experience or not. And
>> that's where the added "bonus" of "not wanting to know how or why it
>> works" comes to bear on the difference between plain shoddy goods, and
>> monopoly crapware. You and your intentional ignorance. Fuck it.
>
>Automatic transmissions outsell manuals by an order of magnatude today.
Yet you see no manufacturer of automatic transmissions plotting to
destroy manufacturers of manual transmissions in the marketplace. Go
figure.
>It's because most people don't WANT to be concerned with the details of
>shifting gears. Likewise, people don't want to be concerned with the
>details of their OS.
You can double-check some of the people all of the time, and you can
double-check all of the people some of the time, but you can only
second-guess all of the people, all of the time. Again, the fact that
manual transmissions are still widely available, and widely purchased,
proves the falsity of your understanding.
>> >Fact: People don't want to become computer literate.
>>
>> Fact: People don't want to do anything but sit on their ass and have sex
>> and watch TV.
>
>And? You're proving my point.
No, I'm mocking your point.
>> Fact: People deal with the real world, where skills have to be learned
>> in order to benefit from them.
>
>Some do. Otherwise everyone would be their own mechanic.
Oh, you read ahead, did you? You'll notice that most people pump their
own gas, a good number of people change their own oil, and there are no
licensing agreements necessary to understand the inner workings of the
engine, if that is your desire. You don't see the auto manufacturer's
changing the specifications of their cars routinely in order to
discourage such behavior, though it wouldn't entirely surprise me if
certain professional mechanics might think that would be a good thing.
>> You don't have to be a mechanic to drive a car. But that doesn't mean
>> driving a car does not require skill and knowledge.
>
>It requires a tiny fraction of the skill and knowledge of being a mechanic.
>You seem to expect everyone to be a mechanic.
On the contrary; you seem to expect that knowing the turn signal from
the accelerator is knowledge only fit for mechanics.
>> >Linux is not that OS (yet). The market *HAS* decided (for now)
>> >that Linux is not yet acceptable for the vast majority of users.
>>
>> The other side of the coin is that it is obvious there is no free
>> market, until it decides there can be a suitable substitute for Windows.
>> That seems to be more reliant on how bad Windows gets, rather than how
>> much Linux might improve.
>
>I don't think so. Windows is improving, not getting worse.
That would explain the wide adoption of Windows ME, and the popularity
of W2K.
>> >> No, we gripe about Windows. That's not FUD; that's being stuck with
>> >> monopoly crapware.
>> >
>> >Stating things that are untrue is FUD, or lying. Take your pick.
>>
>> We gripe about Windows; we don't state things which are untrue.
>
>That's a lie. You yourself have said many untrue things.
Only in your rather bogus opinion, or by mistake.
--
T. Max Devlin
*** The best way to convince another is
to state your case moderately and
accurately. - Benjamin Franklin ***
Sign the petition and keep Deja's archive alive!
http://www2.PetitionOnline.com/dejanews/petition.html
------------------------------
From: [EMAIL PROTECTED] (The Ghost In The Machine)
Subject: Re: Linux, it is great.
Date: Sat, 06 Jan 2001 00:07:20 GMT
In comp.os.linux.advocacy, Aaron R. Kulkis
<[EMAIL PROTECTED]>
wrote
on Thu, 04 Jan 2001 18:29:37 -0500
<[EMAIL PROTECTED]>:
[snip]
>Not only that, but Linux has been ported to IBM S/390.
>
>When, if ever, will LoseDOS be ported to IBM S/390?
www.bochs.com offers an x86 emulator. I don't know what it requires,
but it's theoretically possible that one might be able to run DOS on any
machine that can support bochs' code (IIRC, it's written in C).
Whether this counts is not clear. :-) (Does the S/390 have a floppy drive?)
[.sigsnip]
--
[EMAIL PROTECTED] -- if a machine is emulated by another machine,
how many machines are there and how many Microsoft
Licenses does it take to screw in a light bulb? :-)
up 95 days, 13:19, running Linux.
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list by posting to comp.os.linux.advocacy.
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Advocacy Digest
******************************
