Linux-Advocacy Digest #75, Volume #28            Sat, 29 Jul 00 00:13:04 EDT

Contents:
  Re: MS advert says Win98 13 times less reliable than W2k (R.E.Ballard ( Rex Ballard 
))

----------------------------------------------------------------------------

From: R.E.Ballard ( Rex Ballard ) <[EMAIL PROTECTED]>
Subject: Re: MS advert says Win98 13 times less reliable than W2k
Date: Sat, 29 Jul 2000 03:04:03 GMT

In article <D8vf5.7924$[EMAIL PROTECTED]>,
  "Spud" <[EMAIL PROTECTED]> wrote:
> [snips]
>
> "R.E.Ballard ( Rex Ballard )" <[EMAIL PROTECTED]> wrote in message
> news:8llnb3$plm$[EMAIL PROTECTED]...
>
> > > > Tell you what, why don't you do all
> > > > the work, and I'll collect your paycheck.
> > >
> > > No, thanks, I'm not stupid enough to fall for that.
> >
> > But you ARE!  If you're 45, and you've been working with a PC for
> > 18 years (the first MS-DOS/PC-DOS machine), you've given roughly
> > the equivalent of 3 years salary (adjusted for inflation) to
> > Microsoft related "upgrades".  You replaced machines every 2-3
> > years (avg 2.5),
>
> Which I did in order to run Linux more effectively.

Funny, I was waiting for lower PC prices (like just before Win98
came out, and purchasing Bare-Bones machines for as little as $300
(including Motherboard, Sound, Video, CD, Floppy, and hard drive)
about 1/3 the price of PCs that came out once Win98 was released.

Since Linux is provided by a number of competing providers, and
they don't just assume that you'll buy more memory and hard drive
just to upgrade to the next version of Linux, I've been able to
run Mandrake 7 on a machine I bought 4 years ago.  A little Pentium 90,
with a 1 gig drive.  True, I can't run every single application created
for Linux, but there is enough for the basics, including an OA suite.

>  Which I also did to run OS/2 more effectively.

Actually, I remember running OS/2 on a 486, even Warp 4 ran quite
nicely on a machine that was unable to run Windows NT 3.51 and don't
even think about Windows 4.0

>  Ah, okay, so only a complete and total
> moron would confuse _hardware_ costs with Microsoft-induced costs.

The problem is that Microsoft creates a "feast or famine" environment
in which machines sold shortly after the next release as so scarce,
so expensive, and so difficult to aquire that some companies go on
allocation.  For the year preceding the Windows release, it's "famine",
in which price wars, earnings warnings, and surplusses because no one
wants to get stuck running a machine that's so grossly underpowered
that it can't run the new OS.

Having multiple OS vendors competing for the OS market increases
the number of releases, puts more pressure on vendors to issue smaller
and more managable releases, and to make upgrading from one release
to the next as easy as possible.

> Whoops, you just did that; what does that say about you?  It says you
> have no clue, and are inventing arguments against Microsoft out of
> whole cloth and not even bothering to see if they're _sensible_
> arguments.

Actually, I have only had to make major hardware replacements
(4x previous ram, drive, and CPU speed) for Windows 95 and
Windows NT.  For my Linux machines I buy on the "sweet spot",
just behind that very pricey "biggest, fastest, best" to that
slightly "very good" machine for about 30% off.  Most of this
of course is because the entire supply chain is straining to
keep up with demand for equipment that often requires not only
retooling, but in many cases entirely new facilities (higher density
chips and hard drives require even better filtration, ventilation,
and "noise" reduction.

> Your hatred for Microsoft has addled your mind.  Take a pill.  Take
> several.

I don't hate Microsoft.  They make some very nice products.
I'm just saying that they should have some competition.

At this moment, it looks like Linux is the best candidate.
It's the option most likely to get the agreement of OEMS,
developers, and users.

And no, I don't think Microsoft can create it's own competition.
Even their treatment of Windows 2000 shows that they are essentially
just using their monopoly power to control the marketplace.

If Microsoft were really creating it's own competition, they'd pull
out all the stops and push Windows 2000 as hard as they pushed
Windows 98, price it very aggressively, and let both Windows 98
and Windows 2000 agressively compete with each other for both the
corporate and consumer market.  Instead, they've tried to keep
home users away from Windows 2000, and tried to keep corporations
away from Windows 98.  It hasn't worked that well.  Even with
guarantees of free or extremely cheap upgrades to Windows 2000
(part of the drive to promote NT in 1998 and keep the corporate
users from switching to Windows NT instead of Windows 98), Microsoft
has only sold/given away/generated about 5 million copies in 6 months.
Windows 98 was doing 1 million machines a WEEK by it's 6th month.

Linux is somewhere in the middle.  It's not doing 1 million a week,
but it's doing close to 1 million a month.  Probably because it's
so cheap that some users want to get one of each.

[snip]

> > Of course, you didn't pay it out of your pocket.
> > Some highly placed executive (Director of IT, VP of IT, President
> > of IT, CIO, CTO) made the decision for you
>
> Excuse?  A minute ago, you said _I_ gave my salary over.  Now you're
> saying someone else did this upgrading?  Egads, man, make up your
> mind.

Actually, I've often had the distinct pleasure of sitting in on
the meetings were the decision is made.  Usually, someone who has
been using a pirated copy or beta and holds many shares of Microsoft
stock is pushing for the earliest possible adoption, claiming that
it's better than any other possible choice.

Just to play devil's advocate, I ask if they've done a comparison to
Linux.  In most cases, they come out with some rant like "I'm not
wasting my time evaluating a 1980's text-only operating system written
by a bunch of college drop-outs thats unsupported and unreliable.

What makes this entertaining is that:
 - Linux is upgraded more frequently than Windows
   The code from the recent Linux release is usually newer than windows.
   (only about 5% of the code does date back to the 1980s).

 - Linux supports multiple graphical user interfaces, with superior
   look and feel options to Windows.  The user can actually choose
   between a Windows 9x, Mac, NextStep, Imac, SGI, or Sun/IBM/HP
   look and feel.

 - Linux does contain some code written by college students.  Most
   of these students were BSEEs, MSEEs, or PhDs.  But about 80% of
   the code was written by systems administrators, consultants, and
   the top professionals in the industry.  It was Microsoft that was
   built on code created by a college drop-out.

 - Linux isn't supported by a company that maintains monopoly control
   through copyright, trademark, and trade-secret agreements.  This
   ironically has led to BETTER support for Linux, since vendors must
   compete with each other.  Support providers range from Fortune 500
   companies like IBM to small consulting firms.  Generally, you're
   more likely to get some help at 3:00 A.M. if you choose a larger
   company.

 - Because Linux support is a major point, support providers have
   designed features into Linux that make it more supportable,
   especially remote support.  This includes features introduced
   by SCO and Sun, who have created whole markets based on no-operator
   systems that can be deployed in environments where there are NO
   IT professionals.

> > (you also paid his
> > salary).  The total price - assuming only 6 upgrades, at $5,000
> > each,
>
> Who in their right minds pays $5,000 to upgrade _anything_?  I've
> never done so.  At work, upgrading hardware runs under $300 a year...
> so by your numbers, that $5000 is for over 15 years worth of upgrades!
> Do please, at least *pretend* to make some sense here.

I'd like to see your corporate budget.  The industry average for
a number of corporations is about $5,000 a year.  This was reported
in Gartner reports several times.  Each major upgrade of Windows
has resulted in Gartner Reports and Forrester reports that cite
numbers ranging from $5,000 to $10,000 per user for companies of
over 1,000 employees.

Sure, shoestring companies tend to have people who just come in,
work unpaid overtime, and upgrade 20 machines over a week-end
using drop-in RAM and possibly adding a second hard drive.

In companies of over 1000 employees, we'd be talking 2000 hours
of unpaid overtime.  That's one staff-year.  Even by your best
estimate, it would take 2 hours to do a trivial install/upgrade.

The cost of taking apart and adding SIMMs/DIMMs vs parts replacement
also shifts when you factor in security and anti-theft measures.
Many machines have locks to prevent tampering, and the machine
can only be unlocked when a security officer is present.
Some companies find that things like CPUs, Hard Drives, and
RAM start dissappearing, and it's very often the NEW CPUs,
hard drives, and RAM.

Having 2-3 IT people come in and do an upgrade party on the
weekend for 20-30 machines is no big deal.  Your logistics
consists of ordering in a few free pizzas.  Having 40-50 people
come in and start ripping machines apart is a whole new ballgame.
This takes planning and organization.  The cost actually increses
exponentially, $300 each for 10, $1000 each for 100, $5,000 for 1000,
and $10,000 each for 10,000.  Larger companies try to structure
their upgrades as blocks of 10,000.

>From all of your $300 estimates, I will assume that you work for
a small company that lets you perform your own upgrades on unpaid
overtime.  What about your suppliers?  Are they passing on costs
of a $5000/workstation upgrade to you? (they are if they're over
500 people).

By the way, the URL to my references are in the signature.  Click
the picture in the middle to get to the resumes.  Care to provide
a pointer to your credentials?

My experience includes AT&T, IBM, USWest, Prudential, MetLife,
Great West Life (part of Power Corp), McGraw-Hill, Dow Jones,
FedEx, and the provider of directory assistance for 85% of
(was Computer Consoles, now Nortel Directory Services) the market.

> > And let's not forget those people who work for Microsoft but
> > collect their payroll from your company.
>
> Never met one, actually.

You probably have.  These are people who spend 80% or or more of
their time learning Microsoft products, teaching others to use
Microsoft products, answering questions about Microsoft products,
they have direct line phone numbers and pager access to people at
Microsoft, and regularly participate in the MSDN conference boards.
They are paid by your company, but they seem to have little interest
in the products of companies other than Microsoft unless Microsoft
is promoting them (because they have recently become a Microsoft
subsidiary).

> > You know the ones.
> > The guys who insisted that Windows 3.1 would make a good server
>
> Excuse?  I can't recall _ever_ encountering such a person.

I've seen that at several companies.  At Dow Jones in 1993, their
"official" server was a server system consisting of 3 Windows 3.1
machines.  One parsed and streamed the feed, another loaded the feed
into the database, and another acted as a front-end for clients.
The president of the company had a client to this abomination
providing his personal access to news feeds.  My boss at the time
was the one who insisted that this really was a workable solution.

As a result he was able to get previews of Windows NT 3.5 almost
in the early alpha stage.  As the industry hyped NT, I was sitting
there with my Linux machine running on an 80386 looking at NT
choking in a Pentium 90.  I couldn't help wonder why nobody had
even mentioned Linux or UNIX.  The answer came soon enough.  About
2 days after Walt Mossbarger wrote a nice column on UNIX and Linux,
suggesting that NT would have to compete with Sun quality GUIs,
I mentioned the article in a meeting.  The top ranking executive
at the meeting said "Because of that article, Microsoft pulled
4 full page ads per day from the next 3 days of the WSJ and 2 pages
from Barrons.  We won't be publishing anymore articles about UNIX
anytime soon".  He made a particular point of making sure that I
was present, and looked streight at me.  I was the one who ghosted
the article, and we both knew it.

While I was at McGraw-Hill, I was getting direct information about
the hammering Byte magazine was getting from Microsoft.  Not only
was Microsoft pulling their own ads, they had also obtained the
right to pull ads placed by PC makers and software vendors if the
placement conflicted with Microsoft's interests.  Byte was getting
to the point where it was having a hard time paying the cost of
printing.

At U.S. Clearing, my immediate supervisor, hired the same day I was,
was insisting that all future servers be implemented on NT.  We had
been an SCO shop and I had suggested that we begin migrating to
Solaris.  Adnane decided, after 2 weeks with a copy of Windows 95,
that Windows 95 should be a SERVER for their online trading system.

Every time I tried to point out that Windows 95 was crashing so
often that it wasn't even working as a workstation (this was before
the B release), the VP would agree with me, then Adnane would try
to give me a rediculous assignment with a rediculous deadline (create
a system to clear 1 million checks for which there are no unique keys,
no common fields, and no formal specifications, using free-form printer
formats stored as flat text files - using Powerbuilder and Oracle -
and deliver it in 10 working days.  A comparable project had been
booked for 1 staff-year (4 people for 3 months) and was running late.

I eventually asked the VP if I could do it in PERL.  I delivered
a working system in 15 days.  Adnane attempted to fire me in front
of the VP for not using Powerbuilder and Oracle (I'd used Perl and
Oracle).  The VP stepped in, but I could see that it was only a
matter of time before he got his way.  He started unplugging my
workstation and rebooting it, moving files around, and deleting
my project files from the shared drive (with Windows 95 shares,
there was nothing to prevent him from doing this, and no way to
audit the behavior).  I only figured it out when I realized that
my hard drive had been turned in to shares even when I'd disabled
sharing.  I decided it wasn't worth the effort.  If he wanted to
play games, I'd let him play with somebody else.  I took a better
job for more pay after looking for about 2 weeks.

>  You really *do* work at a totally screwed up office, don't you?

No, I just have about 15 years of experience with Desktop Jockies
who think that because they know Microsoft's products, that Microsoft
is the only workable solution.

> $5000 upgrades. Windows 3.1 as a server.
> No wonder you're so bitter...
> you're surrounded by complete idiots.

The only problem is that these idiots seemed to show up days,
or weeks after my own arrival, and somehow ended up ignoring
all agreements with senior managers in their zeal to turn
my UNIX projects into Microsoft projects.

> Get yourself into a _good_ Windows shop, or even a Linux
> shop, anywhere, as long as you're away from
> those idiots, it's having a deleterious effect on you.

You don't seem to understand.  Until recently, they used
to follow ME.  I'd come in - under the undestanding that I was
going to provide solutions that weren't even within the capabilities
of the existing Microsoft technologies (most would be a stretch
for Windows 2000), and implement them on UNIX, using Linux to
create working prototypes and as a desktop console to larger UNIX
systems (much cheaper than a Solaris or HP/700 Workstation).  About
the time I'd get Linux installed an running, and get access to
the UNIX boxes, I'd suddenly find myself working for a new boss,
who bragged that he owned shares of Microsoft stock that had split
5 times (100 shares of Microsoft stock turned into 3200 shares at
$100/share tends to shift one's judgement).  What was funny was
that each time I suggested that they might do better shifting
to HP or Sun, Microsoft would slump and the UNIX stocks went nuts.

The projects I was working on were things like introducing large
corporations with big brand names to the Internet.  Can you imagine
having a database that gave you access to Dow Jones - running on
Windows NT 3.1?  McGraw-Hill?  Would you trade stock on a 95 server?

In 20/20 hindsight, this seems absurd.  But in 1993, 1995, and 1997,
these were very intelligent men who had been provided with a massive
amount of information sponsored by Microsoft and it's co-op partners.
Meanwhile, I had been directly exposed to real systems which included
both Windows (3.11, NT, 9x) and was able to compare them with real
systems including Linux, Solaris, HP_UX, AIX and SCO.  Even the SCO
systems were running circles around NT in real-world environments,
and I knew it.

In an attempt to persuade me that I was wrong, and didn't know
what I was talking about, I was frequently provided with the best
possible available information about Microsoft's performance
as file servers, group-ware servers, and even as IIS web servers.

One manager proudly dropped a stack of papers on my desk which PROVED
that Windows was 98% reliable.  I told him I had a problem with his
information.  When he asked what the problem was, I told him that
this number was far too LOW, and that he needed to either tighten
up the Windows administration team or change the way he did his
metrics.  I wrote a utility that could be run using the scheduler
that collected statistics and delivered them in a manner that
could be consolidated.  This primarily consisted parsing the alarm
logs for shutdown and reboot indicators.  Later, they installed
real monitoring systems that reported to Tivoli and reported real
availability numbers of 98.9%  They DID eventually tighten the
team and change the way they measured (deducting weekly bounce time
and forced installs.  This boosted availability to 99.8%  When you
excluded the Notes servers, the custom application servers, and
counted only the file/print servers and the IIS servers, the number
came to just over 99.9%

At the time he showed me that report, I did a few look-ups and found
reports showing UNIX availability to be 99.99%, and during the
recalibration period, Sun, HP, and IBM all announnced "5 nines"
meaning 99.999% availability or 10 parts/million down-time.  This
still amounted to just under 1 hour/year excluding 2 15 minute
scheduled reboots (to clear out the zombies).

When Introducing Windows NT administrators to UNIX administration,
I tell them - "The only time you shut down a UNIX box is if it's
on fire".  It isn't that things never go wrong, but you can use
scripts, cron jobs, and services to cause automatic recovery without
resorting to reboots.

> > Just figure that 1/2 your salary goes to feeding the Microsoft
> machine,
>
> Except that not a single person in any organization I've worked for
> meets the descriptions you offer.  If your organization has them,
> find a better organization.

I did, I'm working for a company that is very agressively promoting
Linux and UNIX, both within the organization, as servers, and as
workstations.  They are moving a number of their products to Linux
and have found ways to leverage existing utilities to access their
servers - even though the server applications traditionally ran with
Windows clients.

> > and smile.  You love Microsoft (maybe you're one ofthese guys who
> > works for Microsoft ond company time).  Nice work of you can get it.
>
> I love Microsoft?  Why, because I'm not swayed by blatantly stupid
> arguments "against" it... arguments which often, as you demonstrated
> above, have absolutely no actual basis in reason or in fact?  Sorry,
> wrong door, brain insertion is down the hall.  I don't love Microsoft;
> I don't even particularly _like_ Microsoft.  I do like some of their
> products, but then I also like products from other vendors... whoopee.

The real question is:

If you were offered an alternative to Microsoft that provide more
features, better quality, more flexiility, and more stability, at
1/4 the price - or less, would you be interested?

There is one catch.  You will have to spend 10 hours/week for
3 months (130 hours total) to become familiar with this new
product.  This learning conists primarily of playing with the
different applications, learning some new tools, and asking for
help through provided channels at the first sign of trouble.

This means that you don't spend 100 hours trying to configure
the system before you call the listed support number.  This means
you start the clock once you have a fully configured and working
system (as you would get with any preinstalled system).

This means you spend time trying a few new things each day, not
spend 130 hours using Netscape to browse the net.

If at the end of the 3 months, you find that you are getting good
value, you can still keep you old Microsoft system, but you can
integrate the new system as well.  Eventually, you would have the
option of discontinuing all use of Microsoft and passing on the
next upgrade entirely.

[snip]

> > > What, the UNIX community *is* stupid enough to hand over their
> > > paychecks?  Silly them; maybe they should get some people
> > > with a clue.
> >
> > Look at the intellectual property that Microsoft has expropriated
> > and taken as their own.  The entire Internet was build by UNIX
> > people
> > for UNIX people, with UNIX people creating an inteface that gave
> > Windows users access - now Microsoft is claiming that they had
> > something to do with it.
>
> Something to do with popularizing it, certainly.  Something about
> 100+million installed Windows boxes, allowing end users to use
> friendly and comparatively simple browsers.

But here's the joke.  By the time Microsoft decided to get into the
Internet game, over 40 million copies of Mosaic and Trumpet Winsock
had already been downloaded, Netscape had shipped (electronically
as well as physically) about 10 million Browser/tcpip packages,
and the "Web" was already growing at 20%/month.  In fact, the
web was already having "growing pains" from the exponential growth
before Windows 95 had even been released.

Microsoft was using Internet Explorer to SELL Windows 95, not vice
versa.  Microsoft wanted to prevent Linux and UNIX (Novell's, Sun's
or SCO) from taking over the corporate desktop.  Since most
corporations were dumping "thick client" development in favor of
"thin client/UNIX server" solutions, the Windows workstation was
rapidly becoming irrelevant.  Microsoft used bundling, NDAs,
tie-ins, and lock-out clauses to prevent UNIX & Linux vendors
from cashing in on the hardware that was running 3.1 which was
already obsolete.

Quite simply, Microsoft had 300 million Windows 3.1 machines sitting
on people's desk.  Sun, SCO, Novell, BSDi, Caldera, and Red Hat wanted
to let the user keep that 3.1 machine and run UNIX or Linux on it.
They could upgrade at their leisure.

Microsoft wanted users to destroy all 300 million machines in an
environmentally responsible way, and replace all 300 million machines
with machine that ran preinstalled Windows NT.

Unfortunately, since Windows NT didn't run Windows 3.1 applications
very well (mostly not at all), and since Microsoft wasn't able to get
former allies such as Corel, Novell, IBM, Lotus, Dbase, PeachTree,
Quartedeck, Borland, ... to commit every dollar they had into creating
products than ran exclusively on Windows NT, the OEMs didn't exactly
jump in with both feet.  After the initial "Curiosity buy" triggered
by 4 years of hype followed by a fan-fare release and mass shipments
to thousands of supporters, the NT sales just weren't happening.

Even though Microsoft had a contract with Novell in which Novell had
agreed to stay out of the Workstation market if Microsoft agreed to
stay out of the Server market, Microsoft decided that they could
retain the desktop with 3.1 by announcing "Chicago", at least long
enough to sell a few NT systems as servers.

The biggest problem was that Microsoft was losing Mind-share to
UNIX because more and more large corporations were using UNIX
to do the complex processing and business rules and using Mosaic
compatible browsers to handle workstation display.  By pushing
an NT server, they hoped to keep people focused on the Microsoft
platform.

Of course, with thick-client demand dissappearing and thin-client
becoming more and more popular, there were dozens of managers who
had been managing and developing thick-client projects who found
themselves in very low demand.

This is the reason that there were so many managers trying to push
Windows 3.1, Windows NT 3.5, and Windows 95 as general purpose
server platforms.  All they knew was Microsoft Windows.  They didn't
even know how to estimate a project for UNIX, let alone manage it.

Unfortunately, they also didn't know much about creating servers
either.  They created estimates based on their experience with
Windows Client software, and often found themselves in a crisis,
with projects extremely late (500% on average) and over budget
(800% on average), with no chance of getting through stress testing.

More than once, I found myself being called upon to "rescue" a
project, usually by a very high level manager.  I'd tell him to
order a UNIX machine and by the time it was delivered I'd have
a working prototype running on Linux.  It took me about 2 hours
to move the software to UNIX, get the software up and running, and
have a working system - usually running in PERL.  I'd even set up
UDP servers to exidite the migration to NSAPI - including a very
trivial server that trivialized the migration from CGI to NSAPI.
In many cases, they discovered that there wasn't that much benefit
to NSAPI and often stuck to Apache for everything but "getting the
money".  Most of the time, the executive would then covertly bring
in a couple of UNIX people for the "custom work" on Apache, farm
the "shopping cart" to windows programmers who knew oracle, and
pass the "Glamour pages" (home page, high visibility pages, other
"low risk" pages) to Windows programmers who'd create lots of
"front page extensions" and very quickly pass you off to the UNIX
machines.

>  Oh, sure, similar
> browsers exist on other platforms... but how many of those platforms
> have such a huge installed base... and how many have _had_ a user base
> of comparable size for years now?
>
> > Just like Microsoft, the Internet software was published under a
> > very specific publish license, one that assured there would be
> > no proprietary extensions.
>
> BZZT.  No such license did any such thing.  Try again.

I'll see if I can dig up the NCSA Public License of 1992.  This was
the License under which the NCSA code was produced.  It was almost
identical to the GNU public license.  This meant that all extensions
were to be remanded to NCSA.  Furthermore, NCSA was a federally funded
project.

All Spyglass obtained - and with great reluctance, was the right to
sell "Branding Rights".  They specifically asked for the rights
to change the logo in the corner and title bar.  They also asked
for the rights to preset sights.

When Prodigy attempted to remove the "Location:" Text control
that allowed the user to type in their URL, several contributors
threatened to revoke the license to their code.  Prodigy put the
text control back.  When Microsoft made their purchase, they made
a number of last minute changes to the contract and demanded an
immediate signature (or presumably they'd go to Netscape).  They
signed the deal and realised that Microsoft had included some weasel
clauses which effectively enabled them to make any changes they wanted.
This is when the NCSA unilaterally "took posession" of the NSCA code,
revising the license, and effectively taking ownership.  Since this was
done by the Chicago organization without permission of Cornell (where
the Cello/Windows code originated) or Berkeley (where the Viola/UNIX
code originated), or Marc Andreeson (who had coordinated an written
the dual-platform version), the NCSA was in hot water.  Spyglass was
worthless - essentially a consulting firm not even worth suing.  And
the DOJ stepped in on Netscape's behalf (primarily to cover the
NCSA's butt).

The one good thing that came out of it was that people who were
leaning toward BSD agressively switched to Linux and it's GNU Public
License protection.  This put Richard Stallman in the cat-bird seat.
He eventually created both the legal and technology structure that
put a "wall" between proprietary code such as MQSeries or WordPerfect,
and GNU code such as glibc or Lyx.

The only "pure" Linux is the Debian release, and even the Corel
distribution, which is based on Debian, tacks on some patent protected
GPL, code such as support for GIFs and RSA, on their distribution.

I did get a bit whiney.  I apologize.  I've spent 1000 hours/year
for 10 years cleaning up after windows crashes.  I've spent $1000/year
out of my own pocket to legally obtain Microsoft software required
to read documents created by the clients.  I convert the documents
to RTF and publish my revisions in RTF.

I've also spent more hours than I'd like to think about cleaning up
the messes created because some thick-client windows guru decided
that he could get some of "Billy's Billions" to rub off on him. NOT!

--
Rex Ballard - Open Source Advocate, Internet
I/T Architect, MIS Director
http://www.open4success.com
Linux - 40 million satisfied users worldwide
and growing at over 5%/month! (recalibrated 7/2/00)


Sent via Deja.com http://www.deja.com/
Before you buy.

------------------------------


** 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
******************************

Reply via email to