Linux-Advocacy Digest #719, Volume #27 Sun, 16 Jul 00 18:13:04 EDT
Contents:
Re: Linsux as a desktop platform (Karl Knechtel)
Re: Richard Stallman's Politics (was: Linux is awesome! (Christopher Browne)
Re: which OS is best? (Jens =?iso-8859-1?Q?Pr=FCfer?=)
Re: one step forward, two steps back.. (abraxas)
Re: Linsux as a desktop platform ([EMAIL PROTECTED])
Re: Linsux as a desktop platform (T. Max Devlin)
Re: Linsux as a desktop platform ([EMAIL PROTECTED])
Re: Linsux as a desktop platform (void)
----------------------------------------------------------------------------
Crossposted-To: comp.sys.mac.advocacy,comp.os.ms-windows.advocacy,comp.unix.advocacy
From: [EMAIL PROTECTED] (Karl Knechtel)
Subject: Re: Linsux as a desktop platform
Date: Sun, 16 Jul 2000 20:53:52 GMT
Ray Chason ([EMAIL PROTECTED]) wrote:
: T. Max Devlin <[EMAIL PROTECTED]> wrote:
<snip>
: >Not the scheduling, no, but the weighting, preference, or priority of
: >scheduling. My theory is that with CMT, the market handles whether the
: >end result is valid and useful, and with PMT, it was the engineer who
: >insists CMT is 'stupid' and ridicules people who question that tenet.
: This "theory" has no basis in experience or reason. Indeed, there's an
: enormous counterexample. Windows 3.1 was CMT. And Windows 3.1, from the
: very standpoint of usability and responsiveness you argue from, was a train
: wreck.
Guilt by association.
: Do you remember August 1995? Thousands of people lined up outside CompUSA
: and Egghead and all the rest to buy their copy of Windows 95. Now Win95
: is dodgy and BSOD-prone, but it's nonetheless a vast improvement. PMT is
: a large part of that improvement.
Proof by assertion.
: If you say "preemptive multitasking" to the average Win95 user, you'll
: probably get a blank look. But J. Random User does know that when he
: clicks on "Recalculate," and Excel throws up an hourglass, he can switch
: to Word and type on a memo until Excel is done.
: PMT in action.
I can do this on my CMT Mac just fine (well no, *I* can't, because I don't
use any M$ software. But that's beside the point).
: To get this kind of responsiveness from CMT, you need to break up a CPU-
: bound task. For a spreadsheet, that might mean recalculating some number
: of cells, and then yielding the CPU. It sounds simple, but to do this,
: you have to save your place. Also, the need to break up the task distorts
: the flow of the code; another programmer will have a harder time figuring
: out how it works. You get higher development costs, more bugs, and longer
: time to market. It is unwise to make the developers' job harder than it
: already is.
To get that kind of responsiveness from PMT, you have to let the OS break
up a CPU-bound task for you. It has to save your place and distort the flow
of your code. I hardly see how that's different.
By "flow" of the code I mean CPU utilization over time. If you're talking
about the structure of the source code itself, my understanding is that
doing proper CMT on the Mac is a simple matter of making frequent calls to
the YieldThread() Toolbox routine. I hardly see how that makes for a
significant "distortion".
: I've been there. I've done that. And it ain't no fun. And all too many
: Windows developers said the hell with it and just let their apps throw up
: that hourglass. All too many apps hogged the CPU when they did anything
: non-trivial. *This* is how Windows 3.1 came to be called "Windoze."
I never heard the term "Windoze" applied to any Windows version before
Win95. YMMV.
: You claim that an application that doesn't properly yield will "bomb in
: the marketplace." If that's true for Mac applications, then it is user
: demand for quality that has kept up the quality of Mac apps. PC users seem
: to accept any old POS as long as it has the word "Microsoft" on the
: package.
I agree. How is that an argument in your favour? What's wrong with "user
demand for quality"?
: And user pickiness does not have to go away just because Mac OS X adopts
: PMT. And if PMT could turn a dogpile like Windoze 3.1 into something
: almost usable, imagine what it can do for a Mac.
You still haven't established your claim that PMT is so largely responsible
for Win95's improvement over Win 3.1. Have you ever actually had to *use*
Windows 3.1? I have. Believe me, Win95 has a *lot* of advantages that have
*nothing* to do with internal details like the multitasking system. (I can
hardly believe I just defended Win95.)
Karl Knechtel {:>
da728 at torfree dot net
------------------------------
From: [EMAIL PROTECTED] (Christopher Browne)
Crossposted-To: gnu.misc.discuss
Subject: Re: Richard Stallman's Politics (was: Linux is awesome!
Reply-To: [EMAIL PROTECTED]
Date: Sun, 16 Jul 2000 21:22:33 GMT
Centuries ago, Nostradamus foresaw a time when T. Max Devlin would say:
>Said Peter Seebach in comp.os.linux.advocacy;
>>In article <[EMAIL PROTECTED]>,
>>T. Max Devlin <[EMAIL PROTECTED]> wrote:
>>>Said Peter Seebach in comp.os.linux.advocacy;
>>>>It depends an awful lot on the software. Windows has been sold to tens of
>>>>millions of people, many of whom didn't want it or need it. Niche market
>>>>software often sells for a fixed 25% cut over what it costs to write.
>>
>>>Only for creative values of "what it costs to write it".
>>
>>I dunno. If I have to pay a guy $N/hour to write a piece of software, and
>>I sell it to the one customer for $N*1.25/hour, I think that's a 25% markup.
>
>And what is the markup when you sell it to the next customer, without
>having to re-write it. The customer after that? The thousand later
>customers? The million customers over a three year period?
That is a change of topics.
The topic was _not_ that of a piece of software being sold to a million
customers, but rather of bespoke software written for a niche market
for _ONE CUSTOMER_.
There are many organizations that sell software based on the bespoke
model, where there is _NO_ likelihood of a piece of software being
substantially reused because it was designed specifically to meet the
customer's requirements. That is the matter at hand, and it changing
the question means that the answers change.
>>>Other than getting bought out or having their market disappear, I'm not
>>>sure if very many have. Not the large commercial ones. One product
>>>developer's trying to play the trade secret game die off all the time,
>>>of course, but that's not the same thing.
>>
>>This depends; look at the game industry, where companies fail routinely.
>
>Because you have to continually come up with *new* software to stay in
>business; not just re-sell the old ones over again to the same customers
>as an "upgrade". Most game companies that "fail routinely" were
>startups. We aren't talking startups - we're talking about established
>companies with established products.
>
>>>I'm not sure if you're clear on how "profit" relates to "fixed" and
>>>"variable" costs. The trick with software is its all in the fixed
>>>costs.
>>
>>Most of it, yes.
>>
>>>That's why they're really more of a services business model.
>>
>>In general, yes. Or at least, they should be.
>
>They were, until Microsoft changed the rules, I've heard. I don't know
>the specifics, but I have been told by accountants, who would be the
>ones who know, that the IRS allows software developers to treat fixed
>costs as variable costs. Essentially, they say that the five hundredth
>customer still required $N/hour, because they're still paying
>developers. But the developers are coming up with *new* products, so
>the company is essentially capitalizing on the margin.
Which shows that you don't understand either accounting or taxation.
What the IRS does is of little relevance here; the _point_ of the way
the economics of software works is that the cost of producing the software
generally represents a sunk cost. Once the software is sold once, there
is no additional development cost involved in selling it 100 times or
100,000 times.
The IRS doesn't enter into this in any useful way; when the information
gets recorded in the financial statements, what happens is that they
report aggregate sales, and aggregate costs, and the fact that
reselling software doesn't result in incurring more costs means that
the software developer doesn't get to report additional development
costs. The effect is that if you want to bring in the IRS, the
result is that there's no additional expenses from development,
which, to those that don't understand economics, might elicit the
thought that this _increases_ the tax burden. (It _doesn't_, which
is the whole point of the cossts being fixed.)
"Capitalizing on the margin" is utter business-babble, and
the only place where anybody is going to be pretending fixed costs
are variable is if they are doing internal management accounting
reports, and are mandated to allocate fixed costs as some form of
overhead.
It rather looks like you understand accounting as well as you understand
the CMT and PMT models of multitasking. Hopefully there's some area that
you _are_ expert in that allows you to demonstrate expertise; you have
been doing an excellent job of showing off areas you _aren't_ expert in...
--
[EMAIL PROTECTED] - <http://www.hex.net/~cbbrowne/>
The meta-Turing test counts a thing as intelligent if it seeks to
apply Turing tests to objects of its own creation.
------------------------------
From: Jens =?iso-8859-1?Q?Pr=FCfer?= <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.ms-windows.advocacy,comp.sys.mac.advocacy,comp.os.ms-windows.nt.advocacy,alt.flame.macintosh
Subject: Re: which OS is best?
Date: Sun, 16 Jul 2000 12:46:20 +0200
Bloody Viking wrote:
[...]
> Everyone is different. And yes, there is X Window with Linux. Personally, I
> like the good old fashioned command line interface. My favourite word
> processor is an old fashioned text editor. I use a text editor to code up
> .HTML too, by hand. BUT most people don't want the challenge! I'm not a
> techie, but a long time computer hobbyist. Remember the old Commodore 64? On
> my own, I figured out how to use its BASIC and POKE to code machine code TSRs.
Wow ... you are a Hero! How come people allways try to "prove" their
apparent computer literacy with claims about having poked the C64 or
played around with old VAX back in the dark ages. Who gives a shit! You
don't have to know the wash-board to see the beauty of a washing
machine.
>
> Point n' click is great for routine computer use, like what most people in
> fact do. No question about it. But if for some reason you want to fuck around
> with the computer's innards, a command line is best. I remember Windows 3.11.
> That was fun to fuck with by editing system files or by funny renaming of
> files.
You didn't date much in High School, did you?
> Today, I have one Linux habit with my own box that is like those Windows days.
> I move /usr/bin/vi to something else and rename pico to vi so pico is the
> default text editor, my favourite.
You could of course do it the propper way and simply set the EDITOR
shell variable to /usr/bin/pico if you prefer that p.o.s. editor as
default.
> Like editing International PGP sourcecode
> to remove the "i", I consider this standard practice with Linux. For PGP, you
> chmod +w pgp.c in the source directory tree, edit, and compile. Not hard if
> you know how to use a command line.
This is actually a somewhat lame example of CLI superiority for you can
just as well do that with point and click.
> One thing though. If you own the computer in question, while routine stuff can
> be done by point n' click, you sometimes need to "get under the hood" so you
> still need to learn some command-line skills. No, you don't have to know
> programming, like I know a little of, but knowing some command-line skills is
> a good thing.
No shit.
Why is it I am so fed up with Win-Trolls and lame Linux Advocates? This
Newsgroup seems to be a total waste of bandwidth.
Cheers
Jens
------------------------------
From: [EMAIL PROTECTED] (abraxas)
Subject: Re: one step forward, two steps back..
Date: 16 Jul 2000 21:26:07 GMT
Pete Goodwin <[EMAIL PROTECTED]> wrote:
> I have access to a copy of the final version of Windows Me. Seems to work
> fine with everything I've thrown at it, although they seem to have dropped
> Voodoo drivers from Me. The current Voodoo 3 drivers seem to work just fine
> with it.
>
You must not have thrown a configured TCP/IP stack at it, because theres
known breakage in the irq department. Apparantly thats being 'worked
on'.
=====yttrx
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: comp.sys.mac.advocacy,comp.os.ms-windows.advocacy,comp.unix.advocacy
Subject: Re: Linsux as a desktop platform
Date: Sun, 16 Jul 2000 16:32:37 -0500
On Sun, 16 Jul 2000 20:36:46 GMT, [EMAIL PROTECTED] (Karl Knechtel)
wrote:
>[EMAIL PROTECTED] wrote:
>: On Fri, 14 Jul 2000 16:34:41 -0400, T. Max Devlin <[EMAIL PROTECTED]>
>: wrote:
><snip>
>: If your frontmost task is waiting for human input in a CMT system, the
>: rest of the tasks get very little CPU time, and end up starving.
>: That's the major flaw in CMT.
>
>: If your frontmost task is waiting for human input in a PMT system, the
>: rest of the tasks get the vast, vast majority of the CPU time, and the
>: system continues to run cleanly. That's the major benefit in PMT. A
>
>I highly doubt that.
That's fine. It's wrong, but that's fine. : )
>Unless your computer is psychic, how is it supposed
>to know when will be the next time the foreground app actually *receives*
>input? It won't.
Have you never looked at a graph of CPU utilization, like TOP, on a
Unix or NT box? The system knows what apps are using what CPU time,
and throttles the rest at that priority or lower accordingly.
>And if you don't give very much processor time to the
>foreground app, there's a good chance (because the distribution is
>effectively random) that it won't have processor time at the moment the
>input is received.
Not likely. Perhaps you can provide an example - run RC5 and then
tell me, on a PMT system, that you notice the impact in, say, Agent.
Then run a few more 100% utilization programs, play with the
priorities a bit, and tell me what you need to do to finally notice
it. Hint: You'll need to play with the priorities quite a bit and
either boost the other tasks or lower the Newsreader.
I could point you to a _very_ embarrassing article on RC5 on the
Macintosh platform, where, when put in the background, the Mac's RC5
scores become well....terribly, terribly bad. Would you like to see
that? That demonstrates very well the issues involved in getting
multitasking working on the Mac (namely, that it's very, very
difficult to get even a semblance of "ok", and usually it's downright
terrible.)
>You're going to get either dropped keystrokes or choppy
>responsiveness, unless the timeslice is small (probably not a problem) *and*
>you have a way of buffering that input when the app that has to deal with
>it doesn't have CPU time. I'm not a hardware engineer, but that last bit
>seems to imply (to me anyway) that *something* has to be watching the input
>device on every cycle, possibly an additional processor. (I'm going to go
>further out on a limb, based on assorted things I've heard, and speculate
>that this is how Amiga handles it.)
That's not correct (in the case of the PC or the Amiga...) the Amiga
had one true co-processor (The copper-blitter, I think?) and the rest
of the chips (video, audio, keyboard, etc.) were all just given tasks
by the CPU, but couldn't execute in parallel. In any case, that's not
terribly important to today.
Today, NT/W2k function just fine with 100% utilization - somehow the
OS knows enough to give them CPU time whenever they need it, yet keep,
say, that RC5 client going full throttle in the background. Quite
nice. If you could measure a difference between RC5 in the fg and RC5
in the bg while only running a few other lightweight apps like News
and such, I'd be very impressed and surprised. Not so on the Mac,
where there's an obvious difference in the performance of the client.
>: background render will continue in the bg and will get almost 100% of
>: the CPU time and the user will never know the difference.
>
>: If the frontmost task requires CPU time in a CMT system, it will get
>: it - to the exclusion of (nearly, depending on implementation)
>: everything else. If that's what you want, wonderful. If it isn't,
>
>Not at all. If something in the background also requires CPU time, and both
>apps are written properly, they will constantly yield to each other, and
>thus get as good a split of processor time as they likely would under PMT.
Laughable assertion, destroyed not only by me but also by Znu... How
can they know how often to yield to each other? 1/2 the CPU time, or
1/3? If 3 apps need 100%, they'll do 1/3; what if it's 4 apps that
need 100% time? What if 1 of those apps only needs 1% of CPU time;
how do you manage that? How many different programs and throttling
routines do you want to write for a given application?
>: you don't have the leeway (in MacOS, at least) to do a whole lot to
>: change that. A render put in the bg under MacOS generally dies or
>: gets very poor performance, and the system can become (depending on
>: software) jumpy or 'uneven' as a result.
>
>It slows down POV-Ray for me, but not excessively. On less co-operative
>settings (the Mac POV-Ray is conscious of the CMT environment, and lets you
>'re-nice' the software in its settings) it sometimes jumps to the foreground,
>which is admittedly disconcerting.
>Other apps I have which 'render' things do such tricks as reducing their bit
>depth and dithering in order to reduce their demand on the CPU, conscious of
>the fact that they have been backgrounded.
Are you prepared to say that there is no difference between running a
heavy-rendering job in the BG vs the FG? Until someone can do this,
CMT has a long, long way to go to catch up to PMT.
>: If the frontmost task requires CPU time in a PMT system, it will get
>: it - and everything else that's sleeping will get no CPU time, and the
>: system's idle time indicator (Hi, Chad!) will not increase anymore as
>: the system devotes all CPU time to the active app, and anything else
>: that also required CPU time would "share" the CPU time cleanly;
>: depending on process and priority levels, you might automatically get
>: a 50/50 split, which could be changed at will (I can effectively get a
>
>Again, properly written CMT apps can acheive that split, and exceptionally
>well written ones can allow you to make those changes.
How many are those? Can you show me any? I've never found one that
worked as well as PMT, and I've run a LOT of Mac programs. Face it -
CMT is dead. Its' day (if it ever had one) is over and gone. Move
on. Embrace OS X C.
>: CMT system in W2k/NT by making a fg task 'real-time', which is a
>: misnomer but gets the point across). The user will freely be able
>: to switch between applications with little or no impact to the
>: processes that require CPU time, while also being able to interact
>: freely and without penalty with apps that don't need much CPU time
>: (like, say, a newsreader). That's a MAJOR advantage of PMT.
>
>"Without penalty"? What about all the extra CPU time the OS itself has to
>use up to figure out how to schedule everything?
Compare Amiga's PMT back in 1985 with a 7.14 mhz 68000 to the Mac's
CMT on, say, 68020. Try to tell me the Amiga doesn't destroy the Mac
at multitasking. The "extra CPU time used" (which I think is a myth)
wasn't an issue then, and absolutely isn't an issue now. The people
who use the "extra CPU time used" argument simply don't understand
just how incredibly inefficient a CMT system is.
>The thing about the newsreader you give as an example is that the *average*
>need for CPU time is indeed low, but the *peak* is not. The app sits around
>most of the time, but needs to get several cycles after each keystroke. The
>OS doesn't know when the next keystroke is coming.
And yet, with PMT, it handles the situation very, very well. Yet when
I look at CPU time used by the Newsreader, it's almost zilch.
Amazing, eh? Very efficient. For example, I've been furiously typing
away in Agent for...well, a long time in the past few messages I've
written - probably at least 15 minutes, and how much CPU time have I
used?
49 seconds....Oop! 50 now!
All of the other CPU time? Available to any application that needs
it...almost 100% of the CPU's time is freely available, because even
when I do need CPU time in Agent, it's a tiny, tiny amount.
------------------------------
From: T. Max Devlin <[EMAIL PROTECTED]>
Crossposted-To: comp.sys.mac.advocacy,comp.os.ms-windows.advocacy,comp.unix.advocacy
Subject: Re: Linsux as a desktop platform
Date: Sun, 16 Jul 2000 17:36:05 -0400
Reply-To: [EMAIL PROTECTED]
Said Gary Hallock in comp.os.linux.advocacy;
>"T. Max Devlin" wrote:
>
>> Thank you. That's quite illuminating. Interesting, don't you think,
>> that while many people were saying "the app decides when to yield, not
>> the OS", they probably thought this is just what they were explaining,
>> but none of them thought to put it like that. Perhaps this is why I
>> kept getting flamed for not paying attention when they thought they had
>> answered my questions; none of them realized their answers were
>> misleading. It isn't the OS controlling the multi-tasking which makes
>> the difference; its the notion of a maximum quantum.
>>
>> That makes a lot more sense. Multi-tasking without a maximum quantum;
>> now that *would* be a stupid idea. ;-)
>
>So you finally agree that CMT is a stupid idea for a general purpose OS!
Not *entirely*, no. I was being rhetorical. I didn't think you'd need
to double-check that I knew what I was saying, but I can understand how
you would think you should.
CMT without a convention for limiting a processes' maximum CPU time
*would* be a stupid idea. I'm still not clear on if this is the actual
problem assumed to occur on CMT systems. My post yesterday fantasizing
about a "three level model" for scheduling to replace the monolithic
system found in PMT should indicate that I think CMT/PMT is something of
a false dichotomy. Complete lack of a quantum is certainly an
unacceptable implementation for a general purpose OS, yes.
--
T. Max Devlin
Manager of Research & Educational Services
Managed Services
ELTRAX Technology Services Group
[EMAIL PROTECTED]
-[Opinions expressed are my own; everyone else, including
my employer, has to pay for them, subject to
applicable licensing agreement]-
====== Posted via Newsfeeds.Com, Uncensored Usenet News ======
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
======= Over 80,000 Newsgroups = 16 Different Servers! ======
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: comp.sys.mac.advocacy,comp.os.ms-windows.advocacy,comp.unix.advocacy
Subject: Re: Linsux as a desktop platform
Date: Sun, 16 Jul 2000 16:36:59 -0500
On Sun, 16 Jul 2000 20:53:52 GMT, [EMAIL PROTECTED] (Karl Knechtel)
wrote:
>: If you say "preemptive multitasking" to the average Win95 user, you'll
>: probably get a blank look. But J. Random User does know that when he
>: clicks on "Recalculate," and Excel throws up an hourglass, he can switch
>: to Word and type on a memo until Excel is done.
>
>: PMT in action.
>
>I can do this on my CMT Mac just fine (well no, *I* can't, because I don't
>use any M$ software. But that's beside the point).
Of course you can - if that software lets you switch to other apps.
And don't pretend the original software (doing that calculation) won't
slow down *tremendously* - we both know it will. CMT is a horrid,
horrid system.
>: To get this kind of responsiveness from CMT, you need to break up a CPU-
>: bound task. For a spreadsheet, that might mean recalculating some number
>: of cells, and then yielding the CPU. It sounds simple, but to do this,
>: you have to save your place. Also, the need to break up the task distorts
>: the flow of the code; another programmer will have a harder time figuring
>: out how it works. You get higher development costs, more bugs, and longer
>: time to market. It is unwise to make the developers' job harder than it
>: already is.
>
>To get that kind of responsiveness from PMT, you have to let the OS break
>up a CPU-bound task for you. It has to save your place and distort the flow
>of your code. I hardly see how that's different.
He's talking about programming, so when you say "code" I assume you
are too. And if you are, your argument is, again, wrong. PMT systems
don't require any special code to multitask correctly; the only thing
I'd hope an author would do for a PMT system's software/apps is to
keep CPU usage low. That's it.
>By "flow" of the code I mean CPU utilization over time. If you're talking
>about the structure of the source code itself, my understanding is that
>doing proper CMT on the Mac is a simple matter of making frequent calls to
>the YieldThread() Toolbox routine. I hardly see how that makes for a
>significant "distortion".
Calling other routines, even if absolutely unnecessary? You don't see
how that would cause a lot of waste and inefficiency?
>: I've been there. I've done that. And it ain't no fun. And all too many
>: Windows developers said the hell with it and just let their apps throw up
>: that hourglass. All too many apps hogged the CPU when they did anything
>: non-trivial. *This* is how Windows 3.1 came to be called "Windoze."
>
>I never heard the term "Windoze" applied to any Windows version before
>Win95. YMMV.
>
>: You claim that an application that doesn't properly yield will "bomb in
>: the marketplace." If that's true for Mac applications, then it is user
>: demand for quality that has kept up the quality of Mac apps. PC users seem
>: to accept any old POS as long as it has the word "Microsoft" on the
>: package.
>
>I agree. How is that an argument in your favour? What's wrong with "user
>demand for quality"?
>
>: And user pickiness does not have to go away just because Mac OS X adopts
>: PMT. And if PMT could turn a dogpile like Windoze 3.1 into something
>: almost usable, imagine what it can do for a Mac.
>
>You still haven't established your claim that PMT is so largely responsible
>for Win95's improvement over Win 3.1. Have you ever actually had to *use*
>Windows 3.1? I have. Believe me, Win95 has a *lot* of advantages that have
>*nothing* to do with internal details like the multitasking system. (I can
>hardly believe I just defended Win95.)
True. :)
------------------------------
From: [EMAIL PROTECTED] (void)
Crossposted-To: comp.sys.mac.advocacy,comp.os.ms-windows.advocacy,comp.unix.advocacy
Subject: Re: Linsux as a desktop platform
Date: 16 Jul 2000 21:25:13 GMT
On Sun, 16 Jul 2000 20:36:46 GMT, Karl Knechtel <[EMAIL PROTECTED]> wrote:
>[EMAIL PROTECTED] wrote:
>
>: If your frontmost task is waiting for human input in a PMT system, the
>: rest of the tasks get the vast, vast majority of the CPU time, and the
>: system continues to run cleanly. That's the major benefit in PMT. A
>
>I highly doubt that. Unless your computer is psychic, how is it supposed
>to know when will be the next time the foreground app actually *receives*
>input? It won't.
Good god, it's T. Max's long-lost twin. I wonder how many people who
never heard of blocking I/O are going to pop up with the same bogus
arguments before we can put this thread to bed?
--
Ben
220 go.ahead.make.my.day ESMTP Postfix
------------------------------
** 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 (and comp.os.linux.advocacy) via:
Internet: [EMAIL PROTECTED]
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
******************************
