Linux-Advocacy Digest #760, Volume #25 Wed, 22 Mar 00 21:13:08 EST
Contents:
Re: Why Linux on the desktop? (R.E.Ballard ( Rex Ballard ))
Re: A pox on the penguin? (Linux Virus Epidemic) (Andrew)
Re: Packaging Tools (JEDIDIAH)
Re: Producing Quality Code (mlw)
Re: A pox on the penguin? (Linux Virus Epidemic) (Andrew)
Re: A pox on the penguin? (Linux Virus Epidemic) (JEDIDIAH)
Re: Bsd and Linux (Grant Edwards)
----------------------------------------------------------------------------
From: R.E.Ballard ( Rex Ballard ) <[EMAIL PROTECTED]>
Subject: Re: Why Linux on the desktop?
Date: Thu, 23 Mar 2000 01:09:18 GMT
In article <%kDy4.65$[EMAIL PROTECTED]>,
Bloody Viking <[EMAIL PROTECTED]> wrote:
> John Shields <[EMAIL PROTECTED]> wrote:
> : Ignoring the somewhat "political" issues
> (anti-Monopoly, Open Source, etc),
> : what would be my motivation to run Linux on the desktop?
I've been using Linux since 1992. It's always been a mixed bag.
Since 1993, I've preferred to use Linux whenever possible.
The good stuff:
Linux is relatively cheap.
Although you can buy it free, it's better to pay $30-80
for a supported copy - even today, I often find it useful
to be able to call the distributor - they know the "gotchas"
and can talk me through them in 3 minutes where it might
take 2 hours to fix it the old-fashioned way.
Linux runs on cheap hardware.
I have successfully run Linux on everything from an 80386/16
all the way to a K6-400. The 80386 had 4 meg of RAM, a
dedicated 200 meg swap drive, and a 500 meg root drive.
Applications started slow, but it was functional. It also
worked nicely as a server. Most people preferr a minimum
of a 486DX/50 or better with 32 meg of RAM and a 1 gig drive.
Linux makes a great "thin client".
For customer service and server supported applications, Linux
provides the advantages of a thin client, while still having
enough on the workstation to keep from overloading the network.
Linux is scalable.
As of today, Linux is now the 24th fastest machine. IBM just
announced a cluster of 256 Netfinity servers (4 processors/box)
that cranks over 370 gigaflops and over 370 bips. And it's
all accessible via the X11 capabilities of Linux - one console.
It doubles as an ISP.
When attached to a dedicated line such as a cable-modem, LAN
connection, or DSL line, or when connected to an ISP that offers
"call on demand" ISDN services, Linux can be "standing by" as
a server - no polling for mail, no waiting for someone to "be
home" You can use talk or IRC. You can even use IRC-II with
audio and video since you are providing the server.
I don't have to be home.
I can connect to one workstation from another workstation.
This means that I have the option of accessing a more robust
"home" system via the internet. With the pluggable security
modules, I can have a number of options on how to manage the
connection. In addition, I can encrypt traffic between the
boxes for more security.
I can use Open Standards.
Linux provides a number of tools and applications that encourage
the use of common standards that can be supported by any
workstation including Linux and Windows. This is important
when I'm dealing with organizations which include UNIX people,
Mainframe people, and Windows people. If I publish in Word 2000,
users with Office 97, Office 95, or Office for Windows won't
be able to read, nor will consultants who use SmartSuite, or
government officials who use WordPerfect.
I can add software and remove software without rebooting the box.
This is an especially important feature. Sometimes it's nice
to have software for "ad hoc" uses but you want to remove it
when you need the space for a different project. Linux lets
me customize my workstation to what I'm doing at the moment.
I can manage thousands of documents effectively.
I've been in the industry for 20 years. I've collected
thousands of e-mails, hundreds of documents, and dozens of
projects. When I want to search through all of these articles
for articles that contain desired information (perhaps prior
discovery of a patent-pending idea or concept) I can search
quite effectively using standard UNIX tools.
I can go "Quick and dirty" or "Polished and Beautiful".
Have you ever dealt with someone who wants to spend 3 days
formatting a memo in 25 different fonts that is merely
a request for a $2000 computer? This is behavior that
is encouraged by tools like Microsoft Office. Another good
one is writing a 2 megabyte word document and sending it
with a really descriptive comment like "Here is the document
I promised you". You don't know whether it's an acceptance
agreement, or an Explore-zip virus.
I can use simple tools to than custom software.
In the MS-DOS and Windows world, applications are built from
components that are linked together using a compiler. The
result is a very large monolithic application that has features
you may never use. Linux gives my simple tools like crypt,
compress, uuencode, and sendmail that let me control how
to manage the documents.
Fewer "black boxes"
One of the frustrating phenomenon in the Windows world is when
one black box interferes with another - usually different DLLs
that are loaded based on which application is started first.
Generally the strategy for correcting this is the "4 Rs".
Restart the application
Reboot the system.
Reinstall the application
Reformate the hard drive and rebuild the system.
Often the problem is that a service pack applied to MS-Office
will alter a DLL used by Lotus Notes. Lotus notes will then
"fix" the problem with a service pack that puts the old DLL
in a Lotus directory and specifically loads that DLL. Now,
If you start Word before starting Notes, Notes breaks. If
you start Notes before starting Word, word breaks. Of course,
since the crash usuall happens during race conditions
or deadlocks, it may be hours before the destructive
code is reached.
Fewer Race Conditions and Deadlocks
Because Linux uses services and processes rather than shared
memory and threads, there is less contention for common
resources. The kernel protects the processes from each other,
assures that semaphores are secure, and assures that semaphores
and open files are closed when a process terminates. This means
that fewer parts of the system lock up and there is lower risk
of total system failure.
Error recovery.
When Windows programs access illegal memory or execute illegal
instructions, the behavior can be quite unpleasant. On NT
the application simply crashes, taking application data with it.
On Windows 95, the system may crash entirely. On Windows 3.1,
you can even corrupt the hard drive. When UNIX programs crash,
the kernel "cleans up" and THEN terminates the application, with
or without a "core dump" that can help support people identify
the exact nature of the problem.
Profiling.
UNIX programmers use profiling to find the most frequently
executed code and link that code into common memory pages.
The result is applications that take a bit longer to start
(but don't take over the screen while starting) and run faster
once the initial screen display comes up.
Scripting.
GUIs are wonderful for the initial entry of information for
the very first time, especially when there are strict rules
for what constitutes a valid entry. But when you are trying
to create a daily, weekly, or quarterly summary, scripting
is a very effective way to collect the statistics and summarise
them in a useful format. I don't have to import hundreds of
pages from individual spreadsheets, I can take the CDF documents,
run them through PERL or AWK and summarize statistics from
hundreds or thousands of documents. I can even cull
information from documents containing free-form text and
related descriptions.
Integration with other systems.
Linux can easily be integrated to mainframes and minicomputers
such as corporate databases, market research information, and
business activity - Linux can then summarize massive amounts
of information and generate content that can be used by Linux
or Windows applications to quickly and efficiently create
informative documents. A classic example is taking the
ticker feed from Comstock and cuncurrently computing the
S&P, the DOW, the Russel 2000, and the Welshire 5000 and
graphing all 4 indices before the next quote arrives.
Better load distribution.
Linux encourages systems and applications to be written using
independent processes which are combined using middleware and
clients. The result is a system that can be distributed across
multiple systems when the need arises.
Better storage management.
Linux supports links and symbolic links. This makes it possible
to have the same file appear to be in multiple places at once.
This can be very useful when you want to manage protected and
confidential information.
Better control.
Linux give the user more control over a number of factors which
can impact performance and personal effectiveness. Linux users
can use simple window managers and few colors when they aren't
concerned with "glitz" and are working with minimal hardware,
Better Information Management.
The combination of the features discussed above provide
the ability to manage and leverage available information
more efficiently, making it easier to quickly identify
trends, opportunites, and issues that could impact the
bottom line of the business.
Better flexibility.
Any solution I come up with for Linux can easily be adapted
to any other UNIX variant, Windows NT, Winodows 95, or even
OS/390. Solutions using Microsoft specific solutions can
only run on Microsoft platform. If I implement a solution
in Visual Basic, DCOM, COM+, and MTS, it won't play well
anywhere but NT. If I implement a solution using PERL,
TCL, and/or Java, I can run on any machine that supports
a C compiler and/or JVM.
--
Rex Ballard - Open Source Advocate, Internet
I/T Architect, MIS Director
http://www.open4success.com
Linux - 60 million satisfied users worldwide
and growing at over 1%/week!
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: Andrew <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy,comp.sys.amiga.advocacy
Subject: Re: A pox on the penguin? (Linux Virus Epidemic)
Date: Wed, 22 Mar 2000 20:34:56 -0500
JEDIDIAH wrote:
>
> On Tue, 21 Mar 2000 23:41:48 -0500, Andrew <[EMAIL PROTECTED]> wrote:
> >
> >
> >JEDIDIAH wrote:
> >>
> >> The notion that a GUI
> >> must be big and bloated is a Microsoft phenomenon.
> >
> >And this from a guy that runs X on his Linux box? Care to compare the relative
> >bloat of X and say, Windows 3.1 (a Microsoft product) on a 386 with 4 MB RAM?
>
> I personally wouldn't tolerate Windows 3.1 on 4M. Infact, after having
> bought my first WinDOS 3.1 based PC running on a 486 my first
> subsequent action was to buy an extra 4M of RAM to make the
> responsiveness of Windows barely tolerable.
I wasn't debating the fact that a 386 w. 4MB configuration was ideal for running
a semi-modern GUI. It's not. My point is that X running on a box of this nature
is bloated compared to the Microsoft product. The networking layers, for
example, can definitely be considered bloat.
> I've run X fine on 5M Xterminals actually. Both X and Windows 3.1
> are a dog in low memory configurations. You would be a liar to
> try and imply otherwise.
Believe me I wasn't.
> >
> >> So long as Apple uses Quicktime to effectively |||
> >> make web based video 'Windows only' Club, / | \
> >> Apple is no less monopolistic than Microsoft.
> >
> >This is a very bizzare statement no matter how you look at it.
>
> Not really.
>
> Gutenberg and the printing press had the benefit of a gratis
> open standard. The knowledge of how to decode a book is freely
> disseminated. Similarly, such things as Morse code, ASCII,
> NTSC, teletext, and mpeg are all open. Whereas QT4 is a source
> for vendorlock much like msword where the formats are secret
> and you have to pay a certain select few for access to those
> formats. Those formats aren't even supported genuinely widely
> in binary forms that still protect the 'ownership' of the
> format.
>
> Media on the web presented in RealAudio, Windows Media or QT4
> is no less assinine than converting your web pages over to
> msword rather than HTML. This is especially so when would be
> OS competitors are ignored to their detriminet. It creates a
> market barrier no less problematic than Win32.
Besides the fact that your history's off a bit (literacy wasn't prevalent until
way after Gutenberg and restricted to the few elite), that's not what your .sig
says at all. You're saying that Apple is being a monopolist for, well, not
supporting your platform of choice. QT is hardly a gold standard in web video,
as RealVideo and MPEG, as well as countless new formats cropping up daily, have
a rather significant portion of the market. They're simply not in a monopolist
position. Perhaps one could argue that they're aiding the MS monopoly, but they
certainly can't be considered monopolists in this case.
Andrew
------------------------------
From: [EMAIL PROTECTED] (JEDIDIAH)
Crossposted-To: comp.unix.bsd
Subject: Re: Packaging Tools
Date: Thu, 23 Mar 2000 01:31:15 GMT
On Thu, 23 Mar 2000 01:18:51 GMT, Christopher Browne <[EMAIL PROTECTED]> wrote:
>Centuries ago, Nostradamus foresaw a time when JEDIDIAH would say:
>>On Mon, 20 Mar 2000 02:56:52 GMT, Christopher Browne
><[EMAIL PROTECTED]> wrote:
>>>Centuries ago, Nostradamus foresaw a time when Dr Sinister would say:
>>>>Perhaps you meant to say "Install Shield by contrast is fast and easy."
>>>
>>>Of course, what is more precisely correct is that
>>>
>>> "InstallShield is, in contrast with downloading tarballs, and then
>>> configuring, compiling, and installing them by hand, fast and easy."
>>>
>>>Which all adds up to an overall useless observation.
>>>
>>>It makes sense to compare the use of an RPM or dpkg package with the
>>>use of an InstallShield package.
>>>
>>>Of course, this is a .advocacy group, where comparisons are made when
>>>people feel like making them, where reason plays little role...
>>
>> Even the tarbal vs. rpm argument is useful. Tarballs have the
>> benefit of being much more flexible with respect to dependencies.
>> If you happen to be a minor version behind on some library, a
>> recompile can be simpler. Besides, most tarballs are just a matter
>> of regurgitating a couple of standard sequences.
>>
>> ./configure
>> make
>> make install
>>
>> These are all fairly descriptive and intuitive (given the
>> activity) mnemonics. They should be easy enough to remember
>> after the 5th or 10th time.
>
>A major merit of Ports, dpkg, and RPM are that they can be set up to
>*not require that the human remember anything.*
They still have to remember the package command syntax.
'packages' don't get you away from there. Whereas, the
typical GNU build commands are VERY intuitive. This is
especially true for source rpms.
>
>I can set up a cron job (or a system configuration tool could do
>so...) that goes off and looks for updates from a "trusted" site, and
>which automagically installs things without the poor user even needing
>to know any mnemonics.
>
>If you want to argue for "./configure; make; make install", I suggest
>you bounce that past the people doing BSD Ports work, as they may have
>some elaborations to add...
Anything that's predictable can be automated in a similar
fashion. There's no great handicap in
"./configure; make; make install".
--
So long as Apple uses Quicktime to effectively |||
make web based video 'Windows only' Club, / | \
Apple is no less monopolistic than Microsoft.
Need sane PPP docs? Try penguin.lvcm.com.
------------------------------
From: mlw <[EMAIL PROTECTED]>
Subject: Re: Producing Quality Code
Date: Wed, 22 Mar 2000 20:40:49 -0500
This is a classic argument, in my experience, using 3rd party libraries
is a very expensive proposition. A lot more expensive than the mere cost
of the library, but the balance between learning a 3rd party vendors
library vs writing your own. If the 3rd party library brings a lot to
the table, yes use it, but the third party library MUST be an
unambiguous gain to be justified.
It is almost always better to have core technology developed in-house as
opposed to using a third party library which exposes your project to the
outside forces of a partner which, understandably, may not have your
businesses best interests at heart.
jimmy@free wrote:
>
> In article <[EMAIL PROTECTED]>, mlw says...
>
> >> How many of you actually write a linked list from scratch any more?
> >> how may write hash tables? etc.. if you do, and unless you have a
> >> very good reason to do that, then you are wasting the employer
> >> time and money, becuase instead you should be simply using a
> >> library of those routines.
> >
>
> >This is wrong. Many of the generic libraries available are just that,
> >generic.
>
> NO it is not wrong. You are wrong.
>
> The point is, 99.9% of applications can use those "generic" routines.
>
> Unless you have very specific reason to write your own hash table
> and linked list for 100 time again, use those routines allready written.
>
> It is all a matter of cost/performance. To spend a week debugging
> your new code instead of using code allready written to save one
> nanosecond do not make sense.
Unless it takes a millisecond and you have to run it 300 million times
every couple days for a year. In which case you better put that week of
hard coding in to make sure you have made this the fastest it can be.
That is the issue, tight algorithms can trim milliseconds of each
operation. On large amounts of data, this can add up to hours or days.
I worked at a banking software company. They had a data import function
that took over a day to work on a customers data. We could have lost the
customer. I rewrote the code, eliminated a 3rd party library, wrote the
operations myself. A customer's operation time when from just over a
day, to just over half an hour. Happy customer pays $60,000 in
licensing.
>
> >analyzed a set of log files, over 20 million records in all. I could
> >process it in 5-10 minutes. A guy wrote the equivalent program in perl,
> >it took over two hours to run. In the end, who is wasting time?
> >
>
> You are missing the point. This is not a language issue. It is a
> code reuse issue.
Yes, but it is also a usability issue. If a program takes over two hours
to run vs one that takes 5 minutes to run, the one, used often, which
takes 5 minutes will cost a lot less, regardless of development time,
than the one that takes two hours.
Code reuse is not an issue. 99% of my code is reused one written.
>
> ...
> >sometimes it is better to write your
> >own in a product so there are no royalties or legal issues when you
> >decide to market your product.
> >
>
> Again it depends. If it will $20 to buy the code, and you will
> spend a week to write it, which do you think makes more sense to do?
That $20 code may not be as good as you can write yourself. Using a
generic interface will incur development cost. Importing a foreign
library into a project also requires development work. Seldom is it plug
and play. So, after some amount of development work, plus $20 for the
library, you get a program that will probably not work as optimally as
if you wrote it yourself. You have a third party license agreement with
which to deal.
>
> >>
> >> A good programmer does not have to know all the detailes, as long
> >> as they have the brains to find out about it as needed.
>
> >
> >This is totally false. Being a good developer is understanding the
> >concepts as you write code.
>
> Being a good programmer is knowing where you need to find the
> detailes you need. Actually being a good engineer means that.
> No one can know everything and every detail, but a good engineer
> is good at doing the needed research to find what they need.
True, but programmers that do not know how to program are not
programmers.
>
> >How many times have you needed to sort 3
> >items? How many is the least number of compares required? Can you get it
> >under 3? How many data exchanges?
> >
> >These are classic algorithms that, once learned, affect your thinking
> >and cause you to make better choices as you code.
> >
>
> This is silly. If I am presented with a sort problem I need to work on,
> I have the needed basic skills to look up and research quickly the
> topic at hand and to find the best method to do. I do not have to keep
> all the detailes in my head all the time.
Hash tables, trees, lists, algorithms, etc are what programmers should
know. I would expect a wheat farmer to understand how to grow wheat. I
would expect a chemist to understand chemistry. Why should I not expect
a programmer to understand the very sciences that are programming.
>
> You see, science is not about collecting data, it is about
> the process of finding information.
Yes, one can not do this efficiently unless one knows the basics. I
submit, core algorithms are basic. If you can't do them, work in
marketing.
--
Mohawk Software
Windows 9x, Windows NT, UNIX, Linux. Applications, drivers, support.
Visit http://www.mohawksoft.com
------------------------------
From: Andrew <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy,comp.sys.amiga.advocacy
Subject: Re: A pox on the penguin? (Linux Virus Epidemic)
Date: Wed, 22 Mar 2000 20:40:34 -0500
Mika Yrjola wrote:
>
> Andrew <[EMAIL PROTECTED]> writes:
>
> > And this from a guy that runs X on his Linux box? Care to compare
> > the relative bloat of X and say, Windows 3.1 (a Microsoft product)
> > on a 386 with 4 MB RAM? The relative rendering speed of X vs
> > Windows NT 4.0 (w/o IE4)? Granted, Windows isn't the slimmest GUI on
> > the planet, but just about every modern GUI has its fair share of
> > "bloat".
>
> X is rather bloaty, but a few comments about this in general; X does
> quite a bit more than 3.1, so there is some excuse (although it surely
> isn't as small as it could be).
That was pretty much my point. 'Bloat' is a characteristic of modern general
purpose GUIs, not specifically Microsoft's. I have a lot of respect for what X
is.
> Also, the 4.0 version of xfree86 is
> quite a bit faster than older versions (just try changing from one
> desktop in your window manager to other to see it ;), so it's at least
> in that way better than before.
I'll have to check it out.
Andrew
------------------------------
From: [EMAIL PROTECTED] (JEDIDIAH)
Crossposted-To: comp.os.ms-windows.nt.advocacy,comp.sys.amiga.advocacy
Subject: Re: A pox on the penguin? (Linux Virus Epidemic)
Date: Thu, 23 Mar 2000 01:49:06 GMT
On Wed, 22 Mar 2000 20:34:56 -0500, Andrew <[EMAIL PROTECTED]> wrote:
>
>
>JEDIDIAH wrote:
>>
>> On Tue, 21 Mar 2000 23:41:48 -0500, Andrew <[EMAIL PROTECTED]> wrote:
>> >
>> >
>> >JEDIDIAH wrote:
>> >>
>> >> The notion that a GUI
>> >> must be big and bloated is a Microsoft phenomenon.
>> >
>> >And this from a guy that runs X on his Linux box? Care to compare the relative
>> >bloat of X and say, Windows 3.1 (a Microsoft product) on a 386 with 4 MB RAM?
>>
>> I personally wouldn't tolerate Windows 3.1 on 4M. Infact, after having
>> bought my first WinDOS 3.1 based PC running on a 486 my first
>> subsequent action was to buy an extra 4M of RAM to make the
>> responsiveness of Windows barely tolerable.
>
>I wasn't debating the fact that a 386 w. 4MB configuration was ideal for running
>a semi-modern GUI. It's not. My point is that X running on a box of this nature
>is bloated compared to the Microsoft product. The networking layers, for
NO IT ISN'T.
Win 3.x is just as slow and bloated and nasty as X is.
Some of us wouldn't even tolerate running Windows on that
kind of hardware. I certainly didn't. I thought anything
short of a 486 was a step DOWN from an ST because of just
how painfully slow Windows is below that point.
>example, can definitely be considered bloat.
>
>> I've run X fine on 5M Xterminals actually. Both X and Windows 3.1
>> are a dog in low memory configurations. You would be a liar to
>> try and imply otherwise.
>
>Believe me I wasn't.
>
>> >
>> >> So long as Apple uses Quicktime to effectively |||
>> >> make web based video 'Windows only' Club, / | \
>> >> Apple is no less monopolistic than Microsoft.
>> >
>> >This is a very bizzare statement no matter how you look at it.
>>
>> Not really.
>>
>> Gutenberg and the printing press had the benefit of a gratis
>> open standard. The knowledge of how to decode a book is freely
>> disseminated. Similarly, such things as Morse code, ASCII,
>> NTSC, teletext, and mpeg are all open. Whereas QT4 is a source
>> for vendorlock much like msword where the formats are secret
>> and you have to pay a certain select few for access to those
>> formats. Those formats aren't even supported genuinely widely
>> in binary forms that still protect the 'ownership' of the
>> format.
>>
>> Media on the web presented in RealAudio, Windows Media or QT4
>> is no less assinine than converting your web pages over to
>> msword rather than HTML. This is especially so when would be
>> OS competitors are ignored to their detriminet. It creates a
>> market barrier no less problematic than Win32.
>
>Besides the fact that your history's off a bit (literacy wasn't prevalent until
>way after Gutenberg and restricted to the few elite), that's not what your .sig
While literature wasn't widespread, it certainly wasn't a sealed
secret. What you are trying to attribute to Reformation era
Europe is rather more like chattel slavery in the US south where
literacy amongst the serfs/slaves was just plain ILLEGAL.
Furthermore, there was quite a market for recreational literature
amongst the relatively wealthy urban merchants soon after the
introduction of the printing press.
>says at all. You're saying that Apple is being a monopolist for, well, not
>supporting your platform of choice. QT is hardly a gold standard in web video,
>as RealVideo and MPEG, as well as countless new formats cropping up daily, have
They're a monopolist for encouraging vendor lock. They choose to
compete not based on the relative quality of their product relative
to other choices but by working to ensure that you can't be free to
choose anything else.
>a rather significant portion of the market. They're simply not in a monopolist
>position. Perhaps one could argue that they're aiding the MS monopoly, but they
>certainly can't be considered monopolists in this case.
--
So long as Apple uses Quicktime to effectively |||
make web based video 'Windows only' Club, / | \
Apple is no less monopolistic than Microsoft.
Need sane PPP docs? Try penguin.lvcm.com.
------------------------------
From: grant@nowhere. (Grant Edwards)
Crossposted-To: comp.os.linux.x,comp.os.linux.development.apps
Subject: Re: Bsd and Linux
Date: Thu, 23 Mar 2000 01:55:36 GMT
In article <8bb42m$tr$[EMAIL PROTECTED]>, Brian Langenberger wrote:
>(slightly OT)
Still off topic, but probably accomplishing more than the rest
of the thread...
>You might want to try dvipdfm instead of ghostscript.
I'll have to try that.
>dvipdfm keeps the fonts looking nice onscreen, whereas making
>intermediate Postscript tends to result in jaggy fonts
>(tho the printouts look as good as ever) because of the
>scaled-down bitmapped-ness of them.
I've never noticed any jaggyness as long as I use standard
postscript fonts by doing a \usepackage{times}. Using CMR
fonts does result in on-screen ugliness.
>There's also pdfTeX, but I'd much rather have dvis for
>onscreen previewing.
Me too.
--
Grant Edwards grante Yow! Is this an out-take
at from the "BRADY BUNCH"?
visi.com
------------------------------
** 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
******************************
