On Thu, May 10, 2007 at 01:46:13PM -0400, Brian Gupta wrote:

> AGAIN, SUGGESTIONS ARE WELCOME.

I've suggested External Software Integration more than once, but that
hasn't been popular.

> It's happening. Linux is here to stay. What is being proposed by the

No sensible person expects all GNU/Linux distributions to disappear.
What I am asserting is that faith-based emulation of them is not the
way to build an operating environment that will solve the problems
people care about.

In effect you seem to be asserting that operating environments are
commodities, that they exist primarily to aggregate components
developed outside the community sponsoring the distribution.  This is
antithetical to the value proposition that OpenSolaris strives to
offer: an integrated, *engineered* environment with a single dedicated
community standing behind it.  Consumers of our technology, including
(but not limited to) Sun have found that many end-users understand
that value proposition.  Instead of seeking to emulate the way
GNU/Linux distributions have historically been built, it would be
preferable to understand the spaces in which they have solved problems
more effectively than we have.  This is an opportunity to analyse both
our own shortcomings and those of other players in the space, rather
than blindly repeat what's already been done.  A rigorous analysis of
those problems and a fresh approach to solutions would enable us to
craft a superior innovative answer.  I haven't yet seen evidence of
that in any of this conversation.

To be sure, I agree that there are market segments we're not serving
well, and that better access to the features found in third-party
software would be helpful.  What you're proposing, however, appears to
be nothing but a replay of the approaches that have been taken for the
past 10 years; it offers nothing new.

> top brass at Sun, is exactly what I have been saying. (Personally I
> don't like the fact that we have an operating system named after the
> kernel project leader).
> 
> This is the new market reality. Do what you have to do to get over it,
> and move on. I finally have.

Snide comments like this are not constructive.  OpenSolaris and
Solaris are both having a lot of success in the market today, in some
cases at the expense of GNU/Linux vendors.  Is it possible that you
"got over it" in 2002 or so, when the poor economics of Solaris on
SPARC lost big time to the superior value proposition of fast,
inexpensive x86 gear (which GNU/Linux vendors were well-positioned to
exploit)?  If so, it might be time for another shift in worldview:
OpenSolaris has many innovative features not found elsewhere and, we
believe, superior engineering practices to those employed by the
GNU/Linux communities.  Given our support for the same market-changing
hardware, I think it's fair to say that much has changed in the last 5
years.  Doing what your competition did 5 years ago isn't a winning
strategy, and planning for the future based on trailing indicators
rarely pays off.

> If Sun can attract even a relatively small percentage of the Linux
> community to develop and use OpenSolaris, this push will have been
> more than worth it.

Perhaps you should be proposing this to Sun Marketing instead of an
engineering community.  Our focus here is on building the best
possible operating system using an open, collaborative development
methodology.  Attracting new members to our community is important,
but it's neither the primary goal nor worth sacrificing the integrity
of our work.  Sun may or may not share our priorities, but we've
expressed them consistently from the very beginning.

-- 
Keith M Wesolowski              "Sir, we're surrounded!" 
FishWorks                       "Excellent; we can attack in any direction!" 

Reply via email to