On 03/24/10 11:08 AM, John Plocher wrote:
> On Wed, Mar 24, 2010, Garrett D'Amore wrote to PSARC-ext:
>
>> Ultimately, you won't be able to dictate terms to Oracle about how it should
>> manage its code or its IP, or even its trademarks. At present the ON
>> "upstream" is Oracle's code; the gatekeepers serve as Oracle employees
>> acting in Oracle's interests.
>> ...
>> I don't think anyone (at least in management) at Oracle believes that it
>> is in Oracle's best interest that one of its flagship products should be held
>> completely hostage to the demands of a community who may or may
>> not have Oracle's best interests at heart.
>>
> Wow.
>
> What a complete change of corporate perspective and individual attitude.
>
Not as big a change as you think. Remember the whole opensolaris distro
and naming fiasco? And personally, my opinions haven't changed that
much. I like open source and open development, but I like having a job
with a regular paycheck more.
> If this is indeed a valid reflection of Oracle's mindset, we've gone from
> an open community of developers in 2005 where Sun actively moved
> development and decision making out into the community, to its opposite,
> where everything is slowly getting sucked back into Oracle.
>
Nothing I've said should be taken as official from Oracle. They just
represent my own personal observations. You are free to draw your own
conclusions of course.
> The public signs are troubling:
>
> * A complete silence from Oracle on the future
> of the OpenSolaris developer community.
>
> * ON is seen as "Oracle's code" and not as a
> mix of Oracle and the Community's efforts.
>
Oracle holds the keys to the gate. And largely it holds the code
ownership. The amount of community contributed code to ON is
*miniscule*. The biggest contributions from the community have not been
in code, or docs, but in facilitation of user groups and such. And
ultimately, those efforts don't require anything more than access to the
code.
> * Oracle employees are being actively discouraged
> from playing roles in the OpenSolaris community
> (witness the dearth of nominations of Oracle
> employees in the recent OGB elections)
>
Nobody actively discouraged *me*. I elected not to participate (self
nominate), largely because I don't see how the OGB (regardless of who's
on it) can be effective. Its decisions really carry no weight in any of
my day to day doings. For this same reason, I didn't nominate anyone
else to serve either. I personally think the OGB is likely nothing more
than a figurehead farce and a waste of time -- spending lots of time
arguing about governance issues that really has zero effect on the
actual product that I'm interested in.
These represent my own opinions. Its possible (and I suspect likely),
that given the past history of the OGB's interactions with Sun, others
inside Oracle may have drawn the same conclusions.
> * Significant parts of the Indiana project are finally
> being ARC reviewed, but in private, as closed,
> proprietary cases.
>
I agree that's unfortunate.
> * Our external-to-Oracle OpenSolaris development
> community is being portrayed as an enemy
> that could hold one of Oracle's flagship products
> hostage rather than as a partner who can add value.
>
There is still a way to add value. But leaving the *decisions* in the
hands of the non-employees is probably not going to be something that
can happen.
So far, I think the experiment of deriving value out of the community
has failed. Sun tried, hard. IMO, it failed, and was ultimately
acquired as a result. I don't know the numbers, but I'd be surprised if
the cost to Sun to run the Open Solaris project was not far in excess of
any revenue or other benefit it realized as a result of those efforts.
Its hard (for me at least) to see how market share in a free product
that doesn't have strong tie ins to other profitable products ultimately
leads to revenue. We were losing money on every OpenSolaris CD or
download, but we made it up in volume?
How many people in the community purchased Sun hardware because of
OpenSolaris? I'd wager the number is pretty small, but I don't have any
concrete numbers to back it up. I bet Larry Ellison does, though. (Did
anyone here buy one of the Toshiba laptops preloaded with OpenSolaris?
Did the existence of the preloaded OpenSolaris product influence that
decision?)
> * There has been effectively zero support of the OGB
> in the last year by Sun/Oracle, as evidenced by the
> lack of participation by and interactions with Sun's
> Liaison and the leaders of Sun's OpenSolaris team.
> Combined with a website team that sees itself
> beholden to Sun/Oracle rather than the community,
> and you get an ineffective and powerless OGB.
>
> This sounds more like "proprietary corporate" than "open source
> community" to me. Speaking as a community leader, maybe the Emperor
> really doesn't have any clothes, and it *is* time to make a break - to
> create a "new head of tree" and build a community identity of our own
> out from under Oracle's thumb. Or not - maybe this is an
> overreaction, and we should simply sit back and let things work
> themselves out. In any case, it will make an interesting
> conversation for the new OGB.
>
There is a middle ground between proprietary corportate and open source
community. The code can be open source, and there can be participation
from the community, but don't expect the corporate to place its
needs/desires subservient to those from a community, or spend huge
amounts of money trying to build a community that doesn't tie back into
revenue somehow.
Oracle is *profitable.* I like working for a profitable company.
I like open source too. But the two have to find a balance. I think
under Oracle the balance will swing a bit from where it was.
Ultimately, the thing that *most* benefited the community, and still
benefits the community, is the release of OpenSolaris source code. The
open ARC, open development, and other efforts relating to opening up the
process really provided very little benefit to the community; how many
members actually played any constructive role in any of that?
It would be different if there were a substantial number of external
contributors and contributions. While I think its important that those
folks who do contribute get credit, at the end of the day the cost of
enabling those contributions to Sun was enormous, likely far in excess
of the cost of just hiring a few extra internal developers to do the
work internally.
I'm sorry this post was all about $$ and business matters, but at the
end of the day these issues are really being driven by business realities.
- Garrett
