Today, in celebration of the launch of
OpenSolaris,
I'll take a look back at some
historic events touched by technology and talk about what made them
defining moments. In addition,
two lucky blog readers can qualify to receive a new Sun
workstation, see details at the end of today's blog. At the very least,
today's
OpenSolaris launch
is a defining moment for the software industry and I believe time will
tell it to be
a
historic day as well.
Every industry has its defining moments. For manned space flight, it
was
no doubt the Apollo program that first took man to the moon. I was in
elementary
school then but will always remember that day when Neil Armstrong first
set
foot on the moon. While the first lunar landing was historic in its own
right, it
has become more so simply because man has never returned to the moon
since the end
of the Apollo program, something not even Neil Armstrong could have
foreseen
that day in 1969.
In the 1980's, during the height of the Reagan-era defense buildup,
I
worked for Southern California defense contractor TRW during another
historic
time, helping develop some of the technologies that led to the end of
the
cold war. While many of those programs were similar in scope to Apollo,
TRW
"old-timers" would still recall their contributions to the lunar
landing, including
the first throttle-able rocket engine which landed Neil Armstrong on
the
moon and later, during the ill-fated Apollo13 mission, helped safely
return a
crippled command
module and its crew to earth.
In the early 1980's, the Dec VAX 11-780 was the workhorse of the
defense industry. Even then, however, the writing was on the wall for
Dec.
While from 9 to 5 we ran Dec's VMS operating system on our group's
'780, when most
of the "old timers" went home for the day, I would pull out a 9-track
tape of
a little know (outside academia) operating system called BSD Unix
developed by
Bill Joy at Berkeley
that I had brought with me
from UCLA
and reboot the '780. Along with a few fellow engineers, we would work
late into the night on new software
projects we had dreamed up during the day but not yet had officially
funded by
our customer. Not long afterward, Joy's operating system came to be
used
formally at TRW, this time not on 9-track tape but on some of the first
Sun-1
workstations. Soon, smart program managers who picked Sun as a
development platform
had long lines of engineers volunteering to move to their projects.
Looking back, I'm surprised it took so long, but in January 1994,
after
many Sun development projects at TRW, I finally moved to Sun. SunOS
wasn't just
used for development at TRW anymore, the last project I worked on was
getting
ready to deploy row after row of twenty processor SPARCcenter 2000 SMP
servers
to data centers around the world. What brought me to Sun, however, was
not
its hardware or its operating system, but a new computer language they
had
shown me that was still called Oak
but
soon came to be known as the Java
language. The development of Java technology was clearly a defining and
historic moment for Sun, although no one
involved knew to what extent in 1994.
Fast forward to 2000. While many expected Y2K to be a defining and
historic moment it passed rather uneventfully. With Y2K behind, and the
dot-com
boom in full swing, I was asked to join a new business group at Sun
that would
focus on all of our education and research customers around the world.
The
chance to be technology director for a group that would work with some
of the best researchers and scientists around the world was a great
opportunity.
One of the goals I was given was to help drive the widespread use of
Sun products
and technologies across our education and research customers. I knew
five
years ago that in order to do so, I would have to focus not only on
Sun's
hardware products and technologies, but even more importantly on our
growing
software portfolio.
In order to get as much of Sun's software into the hands of as many
students, professors, and researchers as possible, my group developed
Sun's EduSoft
program, which provides nearly all of Sun's software products to
academic institutions with a no cost
right to
use license for academic and research use. Supporting developers,
especially academic developers, runs
deep in Sun's roots, and in fact the concept of Sun's EduSoft program
was first
hatched in a meeting I had with Jonathan
Schwartz, then VP of Sun's software
group, nearly four years ago. Having started up and run his own
software
company before it was acquired by Sun, Jonathan really understood the
need to support
student developers. However the software world was changing and simply
providing the software to students and developers was not enough. Sun
already was the second largest contributor to Open Source after
UC
Berkeley, but we were not doing enough. More people were running Sun's OpenOffice
productivity suite on Linux than Solaris. So while I wasn't in the
room
the day Jonathan decided to open source Solaris, I soon learned about
the then
secretive project codenamed Tonic.
Timing couldn't have been
better. Sun had already recommitted to supporting Solaris on the x86
architecture and
AMD had publicly announced plans to build a compatible 64 bit processor
now
commonly referred to as x64. Intel would eventually follow suit with
its own x64
designs. Even better, AMD engineers had non-disclosed to Sun on their
now
shipping
plans to
release multi-core x64 CPUs enabling low cost commodity servers of up
to 16
processor cores to be built. What is the significance you might ask?
Hum. Solaris
had been shipping a 64 bit OS for years with scalability across over
one hundred
processor cores. Windows and Linux had little experience beyond 32 bits
and 2 CPUs and no one else had a mainstream x64 operating system.
Better yet,
Red Hat and SuSe, the two leading Linux distributors, were both
starting to
charge what many considered exorbitant prices for their products. But
Sun would
have to convince the community about our intentions not only to make
the
world's best x64 operating system but also that we were committed to
being truly
open. Hence what would become OpenSolaris was born.
OpenSolaris became a huge project at Sun with hundreds of employees
directly or indirectly involved. Kernel engineers worked diligently on
code
cleanup, identifying every last line of encumbered code for Sun's legal
team to
go off and negotiate licenses. About a year ago, the first non-Sun
developers
became involved in a pilot program, which would eventually grow to a
pilot
community of over 140 members. A separate licensing team went to work
examining the
forty plus open
source licenses before eventually creating CDDL with much
input from the pilot community. Like many projects, the effort involved
took
longer than originally expected but the pilot community kept at it,
along with all
the internal teams, while the OpenSolaris site was publicly launched
along
with availability of
Dtrace source code, answering some of the naysayers in the press who
doubted Sun's seriousness to really open source Solaris.
Finally, the time came to plan OpenSolaris' real coming out party.
Should it be on stage in New York City, back home in Menlo Park, or in
some exotic international destination? While the usual marketing
debates examined
the pros and cons of various launch options, someone, I'm not sure who,
had a
great idea. Let's launch OpenSolaris on
blogs.sun.com.
Many OpenSolaris engineers were already Sun bloggers and in fact blogs.sun.com often gets
more traffic than nearly all other sun.com
sites. OpenSolaris has
always been about the community, ever since Bill Joy wrote those first
lines of BSD
Unix or James Gosling wrote emacs, hundreds of other Sun developers
have followed in their footsteps,
contributed to open source projects.
Of course history is yet to be written, but I expect that ten or
twenty years
from now, as some engineer posts to their 3D holographic immersive
communicator
video blog MP30 game player powered by OpenSolaris, they will look back
and
remember June
14, 2005 as a defining moment in the history of open source.
But here is a chance for those of you who don't want to wait twenty
years to get your own pretty cool Sun workstation to run OpenSolaris.
This
workstation is so new, we haven't even announced it, but will soon. To
qualify, here is
all you have to do:
- To qualify, you must be able to identify yourself (by school
email address, listing in school web directory, etc) as a student or
teacher at a
known school
- Download and install OpenSolaris. Then email me
the output of the "uname -a" command from your OpenSolaris system with
the subject "Making OpenSolaris
History", along with information identifying you as a student or
teacher.
- On July 14 2005, I will post a "pass phrase" to my blog. Cut and
paste the pass phrase to your /etc/motd file and email me the output of the
commands "cat /etc/motd" and "uname -a" by 11:59 pm PDT on July 14 with
the subject "My OpenSolaris Defining Moment"
- From the people who successfully complete all steps above, I will
pick a student and a teacher to receive a new Sun workstation, based on
the earliest email timestamp based of the original "Making OpenSolaris
History" email.
This offer is distinct and separate from my other
free
Sun workstation offer of June 6th.
Technorati Tag: OpenSolaris
Technorati Tag: Solaris
