Alan DuBoff wrote On 02/02/07 12:00,:
On Wednesday 31 January 2007 07:09 pm, Jim Grisanzio wrote:
Alan DuBoff wrote On 02/01/07 11:08,:
Let us not forget that the process we speak of is the same process that
made Solaris into the great product that many of us think it is, today.
It's hard to critisize a process that has turned out such a great
product.
An interesting process point in support of this: way back before we did
the pilot, we went around and talked to a bunch of customers, ISVs,
partners, etc. Whenever I spoke with engineers and managers from
enterprise customers about our plans to open up Solaris, they generally
immediately shot back with this: "We love that you are opening the
source, but we don't want you to toss out your development methodology
and all the processes that make Solaris great." Or words to that effect.
This came up in virtually every meeting. But, since we always planned to
/open/ the process, not toss it, we could easily answer this question.
Implementing the change is another matter, I realize, but I also see
that very much on the horizon. Which is cool.
This is really the biggest double edge sword OpenSolaris has faced as a
community. How can we get software putback, but how do we keep the high
standards that have been put in place.
This is the challenge of open development. Alan said it earlier: we did
the open source part, but now we are starting to migrate the project to
open development and that's a bit more complex. Open source is actually
pretty easy in retrospect (though I didn't feel so at the time).
Even the customers were telling us they like the fact that Sun does scrutinize
and hold such a high standard. Since going open with ON, management seems to
want things faster, get things putback faster, and do more.
I doubt things will go that much faster, to be honest. Why should they?
I mean, if a given process sucks and is improved, sure, that will make
that one bit faster. And improving efficiency is always good, and I
think that's a lot of the goal here with the new system we are putting
in place. But in general, I doubt that moving the gate external will
speed things up massively so we can take this flood of new contributions
just waiting to come in. As Rich points out, there are many other issues
that need to be considered and fixed as well. I see things growing
consistently -- even with the gate going external -- in a responsible
way based on solid engineering principles designed by the people who are
closed to this code.
The reality is that if we want to hold the same high standard that has made
Solaris what it is today, we should continue as development has been done,
educate the folks on how things are done inside of Sun, and help them be a
part of the process.
Yes. We are opening our development. Outside developers will learn from
us, and we ought to learn from them as well so that the community gels.
The request sponsor program started this education process and has
resulted in 144 code contributions to date. And a few ARC cases, too.
As with the licensing, let's not point fingers at Sun or the Community, let's
try to work out things the best we can and move forward.
Absolutely. You know, there's a common phrase I often here in open
source: "It will get done when it gets done." Fascinating how that
standard is not applied to Sun. I personally don't really believe in the
statement, but I've come to appreciate that some of these things
/should/ take time and I find myself apologizing for that less and less.
Again, obvious things that need fixing should be fixed and Bonnie's team
has a nifty list of everything now, so I'd expect things to come to
fruition. For my part, I'm telling everyone I meet that community
building takes time -- just as building anything of value takes time. I
think we've all learned something on this project -- both Sun engineers
as well as non-Sun engineers.
Sun has made huge strides in what they've done, and if any of the community
folks don't feel that way, they really are sheltered from the complete
process of how Solaris Engineering works.
_______________________________________________
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org