gnu-misc-discuss
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Too late! Window hasta la vista 5308 is now fully operational.


From: Rex Ballard
Subject: Re: Too late! Window hasta la vista 5308 is now fully operational.
Date: 23 Feb 2006 14:37:11 -0800
User-agent: G2/0.2

Just to keep those words from dissapearing completely:D
Karen Hill wrote:
>
> The lienux kernel of corn rebels had 5 years of opportunity to take the
> desktop  lead.  They failed.  It is too late now.  vista is done, fully
> feature complete.

Actually, there have been those who have been advocating Linux as a
desktop alternative since 1993.  Even in 1993, Linux had most of the
features of a top-of-the-line UNIX workstation such as SparcStation, an
HP-9000, or an RS-6000 or even a Indy.  The screen resolutions of the
*nix workstations were higher, and the monitors were bigger.  Still,
Linux workstations could be produced for under $1,000 while these Unix
workstations cost over $25,000.

Microsoft promised that Windows NT would be a "Better UNIX than UNIX".
They promised that it would have the same multitasking, security,
reliability, and performance as any of the popular Unix Workstations of
the time.  The promise was made in 1992, about the same time that the
work of porting applications such as X11 to Linux first started.

Windows NT 3.1 came out in 1994, and almost died on the vine.  Most
customers who did attempt to deploy Windows NT found that it was
costing as much as $10,000 per users to perform the upgrade - once the
upgrade was done, almost none of the 3rd party applications written for
3.1 would run on the new hardware/software.  The total market
penetration was less than 10% and came far short of Microsoft's
original expectations.

Microsoft then announced "Chicago", the eventually renamed it Windows
95.  Although Windows 95 was officially "released" in August of 1995,
the market didn't really accept or purchase substantial quantities
until 1996.  Windows 95 required Pentium processors, PCI bus, and
peripherals with special "plug-and-play" firmware.  Ironically, this
ultimately lead to the conversion of millions of 80486 machines with
VLB to Linux.  Most early Linux users used the machines at home, or as
"Web Servers" - via dialups connection to the Internet.  It was very
hard to tell the difference between a Linux "client" and a Linux
"server" since X11 servers for Windows and VNC made it possible to run
Linux on a box on the floor, and put Windows on the desktop.  The Linux
machine could be "shared" with several other coworkers.  In many cases,
Linux "servers" were simply hidden in closets in the back corners of
the computer room.  The only thing that distinguished them as Linux
boxes was that there were no keyboards or monitors attached.  Many
Linux servers would be run for months or even years between reboots -
and even then usually only to be upgraded.

Microsoft had to deal with two big threats to Windows 95, one being
Linux, but the even bigger one was Windows 3.1.  People had to be
convinced to migrate to Windows 95, including the purchase of new
hardware which had been secretly designed to NOT run Linux (the PCI
codes required to configure the hardware were a carefully guarded
secret), instead of attempting to upgrade Windows 3.1, failing
miserably, then adopting Linux, OS/2, or UnixWare in desparation.

Microsoft pushed so aggressively that millions of PCs were literally
"dumped on the sidewalks"  - perfecttly functional Windows 3.1 machines
were simply left on public sidewalks or picked up by jobbers who resold
them at computer "shows" or "fairs" which were essentially "flea
markets" for selling used computers.

Many millions of these computers were converted to Linux.  Some were
used as workstations to SUPPLEMENT Windows 95, and others were used as
workgroup servers or as Web Servers for Intranets.  Eventually, many of
these Linux machines, and many of the machines dumped in 1997 were
converted to Linux and shipped to "third world" countries including
India, Africa, People's Republic of China, Eastern Europe, and
Malaysia.  These countries had a combined population of nearly 4
billion people and the market had been virtually ignored by Microsoft.
In some of these countries, they connected systems using 2 meter
repeaters and other UHF repeater and ham radio gear connected to local
area network routers.  Linux had support for both AX.25 and Aloha
networking as well as ethernet.  These were predecessors to the modern
WiFi systems of today.

> The FSF movement has lost most credibility to deliver a desktop
> product.

Actually, OSS has demonstrated that it can be remarkably effective.
The popularity of Apache, PHP, PERL, MySQL, FireFox, and OpenOffice
have shown that the Open Source model works quite well.  Hybrid
products such as Eclipse, Jakarta, Struts, JBOSS, and Xerces - in
commercial implementations such as WebSphere, WebLogic, and SunONE,
have become the foundation of $billions in IT revenue including
servers, software support, and consulting.

This is quite consistent with the business plan suggested by the
original GNU Manifesto, which said that even when software was
available in source form, there would still be a market for packaging,
distribution, support, consulting, and operations management.

> In 2001, they said that their method of producing software
> over then internet with many eyes looking at the code was superiour.
> Could it be true?  The commercial software world was quaking in their
> boots.  Vista was just a gleam in the eye of its creators.  In fact it
> would eventually be rewritten to be based on the Wi n2 k3 code base.

By 2001, the OSS model had already demonstrated it's success.  Many
server applications, along with many desktop applications, including
many third party Windows applications, have been supported OSS
applications which have been made available for download and
distribution.  The availability of inexpensive CD Burners has further
propagated the spread of software in this manner.  In fact, only about
5% of the software available for Windows is still sold through
traditional retail channels, and most of those applications are games,
antivirus, and antispyware - features which SHOULD have been desgined
into Windows from day one.

Many applications now use the cygwin libraries and qt libraries or Java
2, all of which can easily be built for both Linux and Windows - from
the same source code.  Even Microsoft has become concerned over the
lack of acceptance of .net, especially on the desktop.  In fact, the
only really successful Microsoft applications are the "Bundleware"
applications such as Microsoft Office, Internet Explorer, and Outlook.
Microsoft makes good money on their other products, but none of them
are selling in the hundreds of millions of copies.

Conversely, OSS software is being deployed in the hundreds of millions
of copies.  Furthermore, many 3rd party applications have been migrated
to "portable" code, making it much easier to run these applications on
Linux.  It has taken almost 10 years to get the market momentum and the
sustainable growth to bring the ISVs to Linux.  Even then, most of the
offerings are very discreet.  Contracts with Microsoft prevent these
ISVs from making certain types of announcements and packaging choices.
About 80% of the software available for Windows is now also functional
on Linux as well.  The problem is that Microsoft will not allow the
ISVs to advertize this fact, and they are also not allowed to put this
information in their packaging.  Some makers like Electronic Arts, have
provided their "Engine" to Linux distributors.  It's not available in
source code form, so it doesn't ship with Debian distributions, but
they are included with the commercial versions of Red Hat and SUSE.  EA
could very easily include this engine with their games, but something
in the contract with Microsoft prevents them from doing this.

Many UNIX software vendors have also been adopting Linux.  Almost 70%
of the software written for UNIX is now available for Linux and IBM has
developed a "Solaris to Linux" transition kit to make the process even
easier.  Most of the applications that have not been ported are those
which were written when "big-endien" and "little-endian" were not
issues (you'd write for what you had), and ints were 16 bits.  Much of
this code is so old, it's probably not much of an issue that it has not
been ported.

Of course, Linux has also been a hotbed for developers in it's own
right.  Thousands of new applications designed specifically to run on
Linux have been developed.

> The Redmond empire was vulnerable.

It still is.  This is one of the reasons that Microsoft has begun to
diversify into game machines, media outlets, web services, and
investments into hundreds of "partner" businesses.  And even Microsoft
is beginning to make the adjustments necessary to make sure that Linux
and Firefox customers can still make purchases.

The revenues generated by Windows licenses has gradually become a
smaller and smaller part of Microsoft's total revenue picture.  Even if
Vista fell flat on it's face, Microsoft would still be a viable and
profitable company.

>  Anti-trust concent decrees, no new
> os products for 5 years.

And not a huge breakthrough in Vista.  This may be the biggest issue
for Microsoft.  They have been hyping Vista for 5 years, and yet there
isn't an overwhelming argument for an upgrade to Vista - other than
that OEMs will be force-fed Vista.  Microsoft has not cultivated the 64
bit third-party software market, which means that most of the
applications that did run on Windows XP will not run on 64 bit Vista.
Perhaps Microsoft will eventually include VirtualPC as part of Vista,
along with 32 bit XP libraries.  But this would throw the door wide
open to Linux VMs as well.

>  This did not look good, the nightmare senario
> of redmond planners and strategists.  If played correctly by the
> opposition, wills worst fears would come to pass.

Bill did his best though.  You have to admit, getting SCO to create all
this uncertainty over Linux Licenses, threatening to demand $700 per
machine was definitely an interesting scam.  Of course, now it's IBM's
turn to ask some questions, and it looks like the smelly stuff is about
to hit the fan.

>  Yet channel public perception, and enough time could
>  be bought so vista could be born to the world.

Born, yes, but Born to lead - or Stillborn - the market has yet to
decide.
I see lots of eye candy, but I don't see a compelling argument to
upgrade to Vista.

>  Legions of marketers were summoned and deployed.  Linear
> programming, Markov chains,  graph and game theory were deployed to
> calculate the most effective strategies.  Some of the problems were so
> computationally intensive that there are no computers powerful enough
> could calculate the optimal strategies.

Gee, maybe they could have booked some time on Blue GeneL - there are
about 5 really fast computers - all powered by Linux - some of the
fastest in the world. :D

>  Yet when the results were tallied, the FSF is doing WORSE than the bSD
> licensed software projects.  Of note here are the only truly notable
> and remarkable  successful FSF projects:
> gnom
> gCC
> lienux kernel of corn.

It would help if you could spel.

Gnome applications have been quite popular, and the GNOME interface is
quite popular with Red Hat and Debian users.  KDE is prettier, with
more features and eye candy, but at the cost of more memory and CPU
resources.

GCC is still quite popular, with hundreds of applications and
programming languges being implemented using the gcc compiler.
Technically however, it's now the egc compiler.  GCC is very popular in
the *nix community because applications can be compiled to run on
almost any machine capable of running any version of *nix - including
Windows with cygwin or USfW.

Even the Java JVM is compiled with GCC.  Indirectly, GCC is the engine
for most of the applications running on ANY PC.

> That's it.  Most of the other "free" soft wear is done by bSD or giant
> lizard licensed soft wear.  Here is the common soft wear that the FSF
> likes people to accociate with their model yet bSD pre-dates FSF and is
> totally different.

BSD is really the first Open Source license - however, it makes no
distinctions in terms of the derivative products.  There are hundreds
of BSD projects which ended up in the bowels of proprietary projects
including System V Unix and Windows NT/XP.  The irony is that all of
the code that SCO claimed was stolen in public records, has been shown
to be BSD code which was legally available to both Linux and AT&T.  The
same clauses missing in BSD that allow SCO to claim that the code is
theirs, also lets others publish enhanced versions under GPL (Debian).

> A patchy server
You actually spelled it right  - but Madison Avenue likes to call it
Apache

> post a guess
postfix?  Let's not forget xerces, eclipse, jikes, and jbos.

> a nice cup of Joe from the coffe house
Java - Java itself isn't really open source.  Jikes and Blackstone Java
are true open source, and the GCC Java compiler is Open Source, but the
Sun Java compiler and JVM are still proprietary.  It was rather
frustrating when I was looking to install an inhanced Jabber server and
couldn't because it needed Java 1.5 and Sun hadn't ported it to AIX yet
and IBM couldn't do the port at the time.

> and nice open office

What a concept.  "Office Documents" that can be read by people who
don't want to spend hundreds of dollars to purchase, upgrade, and
support special software to read them.

What always makes me nervous about quicken is trying to use newer
versions to generate older tax return information.  I have the records,
but can I read them?  I have to keep boxes of receipts - just in case.

Anyone got something that can read pegasus e-mail?  How about
ButtonWare word processing documents?  And why do my WordPerfect
documents from 1994 look like something out of the twilight zone -
ransom notes?

> etc.
>
> These often get confused with FSF and contribute to making the FSF seem
> like a better way to create soft wear.  Yet, a Fruit Co was able to
> show them this was not the case.

And yet, Apple includes a great deal of open source technology in their
OS/X distributions.  In fact, there was a great deal of resistance and
many developers insisted that X11 needed to be there.  They offered it
as an "add on" - but then they included the X11 APIs in Tiger.

> 2000-2006 will be remembered as the
> years the Fruit Co became successful again in pods to some extent and
> the lienux kernel of corn will just be a footnote in history as an
> unsuccessful idea.

Unlikely.  Google, E-bay, Amazon, E-trade, and hundreds of other
profitable web sites with substantial positive cash-flows have
demonstrated that Linux is a really sound business strategy.  Tivo,
Linksys, D-Link, and WiFi have all made Linux household fixtures and
widely used in business.  Even most of the digital cable tuners have
been Linux boxes.

No, we don't see the little penguin slapped all over these sites,
services, and products, but then again, there isn't some guy with $40
billion in his hip pocket, demanding that his trademarks and logos be
posted - to the excusion of the little penguin, or he will double his
85% markup.

> hasta la Vista will be born at the same time a consent decree expires.
> Gloves come off and the kernel will be eaten for that extra 2% of
> market nutrition.

Don't be so sure about that.  With Microsoft dragging it's feet and
ignoring most of the Judges orders, she might just decide to extend the
judgement another 5 years.  Even if she doesn't, the marketplace has
decided that they can't count on the courts.  Linux distributors like
Novell, RedHat, Mandriva, and Linspire are going right for the jugular,
they are aggressively going after the desktop as the primary desktop
operating system.  Companies like Dell and HP have been very clear that
if they want to be able to offer desktop and laptop machines with 64
bit processors and >4 gig memory and terabyte storage systems - they
need to go with Linux.

In fact, Microsoft may end up being an "application" under Linux - the
OEMs can offer crossover or install Windows XP in a bochs.  Since 64
bit Vista doesn't have enough applications to "lock in" the market -
It's quite likely that even Vista will end up being put in a Bochs.

> By the way, My Nathans brilliant plan to put  Eunuchs  out of the top
> spot market share in the serving business worked finally after more
> than a decade.  Bet you read about the story in the news! People prefer
> a window to serve with than a Eunuchs now.

Yes.  Unix servers got more and more effecient.  Fewer were needed,
they could do more with less resources, and they were cheaper to
maintain.  As a result, little single-task Windows servers, which can
only run a single component per server, and require hot-standby units
to maintain availability - have become more numerous - and more
expensive than both UNIX and Linux servers.  Let's face it, I can take
one Z-9 server and do the work of about 500 Windows servers using Linux
and OS/390 VMs.

>  How did they do it?  By
> convincing supporters of for pro fit Eunechs creators to go with a free
> lienux.  Just look at the 1999> credits.  Grep them some time.

Amazing isn't it.  HP has a Linux that is really secure and fast.  IBM
has Linux compatibility on AIX and has Linux running on X-Series,
Blades, P-Series, I-Series, and Z-Series.  HP can run Linux on
Superdomes and has Linux compatibility on HP_UX.  Sun has StarFire
servers that can run Linux and Solaris - at the same time, on the same
machine.

Poor little Windows 2000 can only run a single function, like a
database, or a web server, or a Web Service Engine, because of all the
DLL hell conflicts.  Windows 2003 can run multiple virtual machines,
but even then its' mostly for "terminal servers" and only 3 "terminals"
are included with the service.  Linux offeres unlimited GUI level
access.

And of course, since Windows servers need to be rebooted, tend to hang
for no good reason, and require manual configuration using GUI
interfaces - you need to run them in pairs, with one doing nothing
while the other does all the work - just in case something goes wrong.

> Less biz for the pro fit eunuchs means less RD and less competitive pro
> ducts tape.  Aye eye Ex, hype You Ex, So Lier is, were food munchies
> lienux ate just like Nathan planned.

I think the big IT budget was being gobbled up by Microsoft.  Prior to
2000, billions had to be spent trying to fix all the Y2K problems.
Then there were all those upgrades.  And most of these upgrades to
software also involved hardware costs and labor costs.  In many cases,
the cost was as much as $2 million/year/thousand users, an average of
$2,000/year - roughly 1 months pay for the average office worker.

The Windows server was also a mess.  Many companies would put projects
on Windows thinking they would only need one or two servers, and
suddenly found themselves needing thousands of servers to handle the
loads of a large organization in even the most basic applications.
Many companies gradually migrated their Windows NT 4.0 applications to
Solaris, AIX, or Linux - using Windows 2003 as Front-Ends for
compatibility with legacy NT applications.

Microsoft decided to triple the support prices for Windows XP and there
were STILL the hardware and labor costs of the upgrades even.  For many
companies this was the final straw.  They signed the contracts, most of
which expire this year, but they also began to create formal migration
plans, implemented over a 3 year period, which would allow them to
quickly transition to Linux in a matter of weeks.  This would at least
give them more bargaining leverage for the 2006 and 2007 negotiations.

This is the market which Vista is entering:

  Instead of a market where there are millions of computers which can
run nothing but Windows, there are millions of computers computers that
could be converted to Linux very quickly.

Instead of a market where the only software capable of reading
documents runs exclusively on Windows, a freely downloadable
application (Open Office) can provide access to these documents with
very little effort.

Instead of a market where Microsoft hold all the cards and can force
OEMs and Corporate customers to configure the system to EXCLUDE Linux,
Linux is currently just another application which gets kicked off and
run from VMWare player.  Microsoft will have to negotiate just to be
sure that it doesn't end up stuck in a bochs engine with minimal amount
of memory.  Even worse, the OEMs and Corporate customers might not even
WANT Vista - they might actually decide that they want to stay with
Windows XP - as a secondary operating system.  They've already paid for
the licenses.  If Microsoft tries to Force the issue, it will simply
accelerate the migration to Linux.  Keep in mind that one of the
problems Microsoft faced when they tried to release Windows NT 3.x and
Windows 95 was that there were so many millions of Windows 3.1 machines
and so much Windows 3.1 software - and people weren't willing to just
jump into Windows NT just because Microsoft told them to.

It's a big huge gamble for Microsoft.  Microsoft has bet the ranch 4
times - and lost.  When they released Windows NT 3.x, it bombed, and
they almost lost control of the market.  When they released Windows NT
4.0 the market acceptance was painfully slow, they had to throw in NT
5.0 (Windows 2000) for free as part of the package to get acceptance as
a corporate workstation.  And Windows ME was such a disaster that
Microsoft couldn't replace it fast enough.  Customers who purchased ME
often demanded "downgrades" to Windows 98, and many ME machines were
converted to Win4Lin or Linux/Crossover machines because ME was so
disliked.

Microsoft took no chances with XP.  They demanded that corporate
customers immediately accept shipments of millions of XP licenses, even
though they wouldn't be deployed - in many cases for over a year.
Microsoft also tripled the support prices, telling corporate customers
that if they didn't sign up today, they would be on their own (even
though OEM customers got the same updates and service and support as
the corporate customers.

I'm sure Microsoft has been thinking all of this out very carefully.
I'm sure they have a few "dirty tricks" up their sleeves, and I'm sure
that they think that they can bully the market into accepting a "force
feeding" of Vista.

>  But then a big blue meanie  took
> the initiative and began feeding  lienux Aye eye Ex and turned the
> tables on My. Nathan.

Linux and AIX are two very different critters.  About the only thing
IBM gave Linux was the schedular they had developed for VM/CMS and
OS/360.  It works really well, and it's much simpler than trying to
manage hundreds of semaphores, or one big fat one.



reply via email to

[Prev in Thread] Current Thread [Next in Thread]