Home Messages Index
[Date Prev][Date Next][Thread Prev][Thread Next]
Author IndexDate IndexThread Index

Re: Microsoft Has Stopped Competing with Linux

__/ [ Rex Ballard ] on Friday 24 March 2006 06:22 \__

> IBM really isn't in the OS business, or even in the business of Linux.
> 
> About the only OS the actually "Own" exclusively are the "mainframe" OS
> modules such as ZVM, VM/CMS, and MVS.
> 
> They are licenced to offer UNIX (AIX) to which they have added
> enhancements from these other Mainframe operating systems.
> 
> They are licensed to offer Red Hat Linux and SUSE Linux, to which they
> have added enhancments from the Mainframe operating systems.
> 
> IBM even offered these Mainframe OS enhancements to SCO for their
> attempt to port a 64 bit version of SCO Unix.
> 
> The irony is that SCO decided that they wanted to sue IBM for licensing
> the little bit of code IBM owns exclusively, to Linux.
> 
> Microsoft's competition is not IBM, it's Red Hat, SUSE, and the other
> major commercial Linux distributions.
> 
> IBM is simply a satisfied customer who uses the Linux operating systems
> of the two competitors as a means of adding value - as a VAR.  In fact,
> IBM doesn't even make their own PCs anymore.  They purchase them from
> OEM Lenovo, with OEM Microsoft licenses, and then as a Value Added
> Reseller, add features such as Lotus Workplace, and even Linux (for
> it's IGS consultants only) Client for E-Business (C4EB).
> 
> Microsoft is not happy that IBM has invested over $1 billion in Linux,
> and they are even less happy that IBM now makes over $16 billion/year
> in revenues generated by Linux opportunities, not including Services
> engagements (since most IGS engagements aren't "pure" Linux
> engagements).
> 
> In 1997, suggesting Linux as even a gateway was a risky proposition.
> Most CIOs and upper IT managers had gone directly from Mainframes to
> Windows, and really didn't know much about UNIX.  They knew that UNIX
> systems were cheaper than mainframes, and that it didn't take a big
> team to support a big UNIX server farm, but they had no "Hands On"
> experience with Unix.  As a result, when Microsoft suggested that they
> should replace their mainframe and Unix systems with Windows NT 4.0
> SP1, many of these executives thought that this was a good idea.
> 
> The problem was that things didn't go as planned.  Many larger projects
> never got off the ground and had to be reimplemented on Solaris.  Those
> projects that did survive were mostly very simple projects, such as
> very small databases for a small organization within the company, mail
> servers for 10-50 people, or web servers used as "front-ends" to UNIX
> enterprise databases.  The problem was that in a large organization, it
> wasn't unusual to need over 100 servers per thousand people.  Some
> companies had 3 to 4 THOUSAND NT 4.0 servers.
> 
> The problem was that these servers needed to be rebooted at least once
> a week.  In addition, they often needed rebooting midweek.  They needed
> one person for every 10 servers for every shift.
> 
> Linux servers were usually introduced by the UNIX administrators, and
> these PCs were often "hidden" in the back, often in a closet or on the
> bottom of a rack, connected to nothing but an ethernet cable.  Very
> often, they were used for e-mail, web servers, file sharing, and even
> as routers and programmable firewalls.
> 
> In fact, in 1996, most CIOs didn't even realize that Linux was being
> used, and in 1997, 17% of those surveyed found that they had not only
> been using Linux, but often quite heavily.
> 
> The "Little Sister" of Linux was FreeBSD, which also gained
> extraordinary acceptance, especially for use in
> single-processor/single-server desktop-turned-server situations.
> 
> Last year, Microsoft finally declared that Windows NT 4.0 was
> "obsolete" and refused to even issue security updates.  They approached
> all of their NT 4.0 users and insisted that they upgrade to Windows
> 2000 or Windows 2003.  The only problem was that there were differences
> between the licenses and prices of these newer versions.  Licenses for
> NT 4.0 were typically selling for around $1500 per CPU.  For "Data
> Center" servers, Microsoft wanted as much as $25,000 per CPU.
> 
> Most of the customers approached had already made their plans for a
> hasty migration to Linux.  In some cases, they were even able to use
> Linux with WINE to run the pieces they couldn't port.  More often,
> vendors were quite eager to offer Linux versions of the popular
> applications such as Oracle, Sybase, Seibel, SAP, and many others.  The
> irony is that it often cost less to port from Windows to Linux than it
> did to keep the boxes running for a year (partly because all of the
> Open Source accelerators available).
> 
> The irony is that as these NT 4.0 machines were being replaced with
> Linux, those who trained themselves to do the scripting and automations
> used to keep Linux "self maintaining" as much as possible - found that
> they had time to support development efforts - staff that had primarily
> been focused on "box booting" were now active participants of strategic
> projects and initiatives.
> 
> Nearly all of those graduating from college in the last 2-3 years have
> been quite proficient in Linux, C++, scripting languages such as PERL,
> Python, and Java.

While on the issue, some graduates are whining that they are not equipped
with skills that are vendor-bound. For example, some graduate find that .NET
has some value in industry. As previously reported, certain courses finally
discourage the use of packages like Photoshop and use OSS instead. What it
boils down to is the need to teach *principles* rather than pricey tools
that make the course seem like industry-funded preparation for a life-long
(make believe) lock-in/relationship.

Sorry I could not keep up with your lengthy post. As a follow-up to a fairly
comprehensive overview, I fear that our role as advocates is becoming less
necessary. Linux spreads itself already and it gained the necessary
momentum. It's mainly a matter of giving it a more considerable thrust now,
making the UNISYS and Apache's Stein prediction (10 years to death of
proprietary software) come true even sooner.

Best wishes,

Roy

-- 
Roy S. Schestowitz      |    Othello for free: http://othellomaster.com
http://Schestowitz.com  |    SuSE Linux    ¦     PGP-Key: 0x74572E8E
  9:25am  up 16 days  2:02,  8 users,  load average: 0.75, 0.67, 0.84
      http://iuron.com - next generation of search paradigms

[Date Prev][Date Next][Thread Prev][Thread Next]
Author IndexDate IndexThread Index