On Jul 3, 10:52 pm, GPS <georg....TakeThisOut@xmission.com> wrote:
> "So I'm pretty well-known for not exactly being a huge fan of the FSF and
> Richard Stallman, despite the fact that I obviously love the GPLv2 and use
> it as the license for all my projects that I care about."
> ...
Linus originally released Linux under the GPL because he never planned
for it to be anything more than an interesting curiosity, and possibly
something that the GNU HURD project could borrow from. Linus was a
graduate student in Finland, and had written Linux as a study in using
an MMU. He had taken a course in operating system theory based on
Tanenbaum's Minux operating system, but realized very quickly that if
he really fully utilized the MMQ of the 80386, that he could make a
much simpler operating system that would also be significantly faster
than Minux. Minux stood for "Minux Is Not UniX" because although it
was very loosely based on the fundamental concepts of BSD Unix, and
Version 6 Unix, it did not support memory mapping, memory protection,
segmentation, and was designed to be implemented in an 8086 or 80286
processor.
Linus didn't even call his brainchild Linux, he just said something
like "I wrote this little operating system, try it out and let me know
what you think". It was only 10,000 lines of code, but it was more
than enough to impress hundreds of other developers who were so
inspired that they began suggesting an number of improvements,
enhancements, and ways to make it more API compatible with various
flavors of UNIX including SunOS and BSD Unix. Soon various BSD and
GNU applications were ported to what had been finally named "Linux", a
play on both Linus and Minux.
> "This is, just to take an example, one of the reasons I try to avoid talking
> much about Microsoft - I'm very passionate about Linux (obviously), but
> quite frankly, I really find the whole notion of Linux as being "against
> Microsoft" to be silly and wrong-headed ..."
Linus didn't actually "promote" Linux. He continued his studies for
two more years, during which more and more contributors to GNU, HURD,
and disgruntled contributors to BSD who were upset that AT&T had
acquired their code without giving them any form of formal recognition
or royalties. Most of the X11 software was ported to Linux
backwards. Rather than altering the X11 code to run on Linux, Linux
contributors would make minor tweaks and changes to the Linux code and
library code to get the X11 code and other GNU and BSD application to
run on the Linux kernel. The military had set up a huge repository of
Open Source, Public Domain, and other "unrestricted" software, most of
which had been ported to Unix, available on their SIMTEL20 server. By
the beginning of 1993, nearly all of the available Open Source
software was running on Linux because it was now so API compatible
with BSD Unix.
One of the first commercial distributors of Linux was Soft Landing
Systems in Toronto Canada. They offered the software for free
download via BBS boards and anonymous FTP servers, but they also sold
an "Ammo Box" of 100 floppy disks, organized so that you could start
with a system as small as two floppies to try it out, but could create
a partition or install a second drive on a Windows 3.1 or Windows 3.11
WFW PC and configure it with up to 100 megabytes of software. SLS
Linux also supported SCSI drives and CD-ROM drives. By April of 1993,
SLS was offering a CD-ROM containing all 100 "Packages".
By June of 1993, there was SlackWare Linux out of Phoenix Arizona,
from Patrick Volkerding. In October, the first Red Hat Linux
"Halloween" release came out. Linux was almost shocked by the success
of these "Linux Distributions". Even at this point, he was still in
College, but by May of 1994, the Red Hat "Mother's Day" release, which
featured "Plug-and-play" functionality which could detect the
configuration of such things as IDE drives, CD-ROMs, Floppies, SCSI
drives, and Network devices, and automatically configure them. What
made this even more notable is that this "Probe" technology worked on
ISA, EISA, and VESA bus computers. Within a few months, it also
worked on Micro-Channel devices. The use of loadable "modules" was a
radical departure from previous UNIX implementations which had to be
recompiled or re-linked with the appropriate drivers. In effect, you
needed to create a "custom kernel" for each PC. The Linux "modules"
approach used shared libraries that could be loaded and called by the
kernel. More importantly, it was possible to call proprietary code
from Modules without releasing the modules under GPL.
Sometime between Mother's Day and Halloween 1994, Microsoft became
acutely aware that not only was NT not meeting it's sales target, but
Linux was now threatening to make it possible to run Linux on millions
of Windows computers that were not capable of running NT. Linux could
run on 4 Mhz 80386 machines and featured full X11 graphics mode. It
could run with as little as 4 megabytes of RAM compared to the minimum
32 meg required by NT 3.x. At the same time, when installed on high-
end PCs such as 80486-DX/100 and Pentium machines, especially machines
with 8 or 16 megabytes, Linux could outperform most $35,000 UNIX
Workstations, even though it didn't have support for commercial
applications yet.
The rise of Linux caused an even bigger problem for Microsoft. Many
corporations had stopped upgrading to Windows NT, Novell had purchased
Unix and was about to release a PC workstation version called
UnixWare, and IBM's OS/2 WARP was now stable enough to be attractive.
The result was that Windows NT, which had fallen far short of
expectations, was faltering, Gates had promised "Chicago" but couldn't
make it just a trivial upgrade to Windows, and Red Hat, Novell, and
IBM were all offering their Operating Systems on a non-exclusive basis
for a fraction of the price Microsoft was charging for Windows, if the
OEMs pre-installed them in a "dual boot" partitioned system, allowing
the end-user to choose which system they wanted to use each time they
booted the PC.
What followed was a very intense campaign for what could have been
Microsoft's very survival. What could have become, in 1995, a
unification around UNIX and Linux as the new standard for desktop PCs,
was stopped by Microsoft, using a combination of fraud, extortion,
blackmail, sabotage, and obstruction of justice.
When called by the defense at the DOJ vs Microsoft trials, first in
relation to the violation of the 1993 settlement, contempt of court,
and then the Antitrust cases, Microsoft executives, especially Bill
Gates, admitted to these criminal acts, but not using these exact
words, arguing that these were a form of "Corporate Self Defense". In
effect, Gates was claiming that although what he did may have been
unethical, immoral, and possibly even illegal, he was justified in
taking these actions in much the same way that a father could kill a
perpetrator about to kill him or his children.
In Judge Jackson's "Findings of Fact" he very specifically pointed out
that the prosecution had made no attempt to prove that the Monopoly
had been obtained illegally, but Jackson also pointed out in his
findings of law, that Microsoft used it's existing monopoly power
illegally.
> http://torvalds-family.blogspot.com/2008/11/black-and-white.html
> I am impressed by Linus' understanding, and insight. I may have misjudged
> some of his actions.
Actually, Linus was very wise to stay detached from the cases against
Microsoft. To begin with, he had been on Paul Allen's payroll -
perhaps because Paul Allen often invested in top competitors to
Microsoft, including AOL, and Linus' employer Transmeta. Allen's one
request was that Linus implement a version of Linux that ran better on
the Transmeta chip than on traditional Intel Pentium PCs of the day.
Ironically, many of the best features of the Transmeta chip, RISC
chips, and other high end chips began to trickle into the Intel chips,
not to support Windows, but to support Linux. By 2003, most
corporations and government organizations had informally established
to purchase only hardware capable of running Linux and running it
well. They had been forced to upgrade to Windows XP and the SP2
release had rendered numerous applications inoperable, requiring a
very costly roll-back and blockade to prevent Microsoft's automatic
upgrades from shutting down their corporate communications and
collaboration.
The release of Vista only made these companies even more wary of
Microsoft's anti-competitve tactics. Vista blocked the installation
of many 3rd party applications, and failed to run many others. The
result was that Corporate customers insisted that PCs support both
Windows XP and Linux as well as Vista. In addition, many corporations
and organizations are now turning to desktop virtualization to assure
that Linux systems can run XP as VMs even if the PC won't support XP
in native mode. Even Microsoft is becoming acutely aware of this
demand for virtualization and has introduced virtualization on their
Windows 2008 servers as well as possibly introducing it as part of
Windows 7.
By avoiding direct "Microsoft OR Linux" and instead maintaining a
strategy of "playing nice" with Windows, Linux has spread to hundreds
of millions of PCs, appliances, and other virtual and real devices.
It's possible that the total number of Linux/UNIX deployments now
exceeds the total number of Windows PC deployments.
> -GPS
Hot
Bloated Email? DetachPipe for Outlook
Mac
Linux
Games
PDA
Mobile
FAQ
Search
Profile
Private Messages
Log in
Linux Forums
RSS