OpenJDK Governing Board Minutes: 20011/4/21

[Mike Milinkovich]

Andrew,

This unhappy situation will hopefully last for at most a week or two. If it appears that it will be putting open development at a serious disadvantage, we will certainly review the decision.


> -----Original Message-----
> From: discuss-bounces at openjdk.java.net [mailto:discuss-
> bounces at openjdk.java.net] On Behalf Of Dr Andrew John Hughes
> Sent: May-03-11 4:02 AM
> To: discuss at openjdk.java.net
> Subject: Re: OpenJDK Governing Board Minutes: 20011/4/21
> 
> On 30 April 2011 11:59, Doug Lea <dl at cs.oswego.edu> wrote:
> > On 04/28/11 15:06, Ludwig, Mark wrote:
> >>
> >> It might help the contributors be more patient if they understood why it
> >> helps the OpenJDK community to wait for this.  Neither the minutes from
> >> the
> >> board meeting nor the ensuing discussion in the last twelve hours really
> >> explain why waiting indefinitely is better than letting OpenJDK 8 get
> >> started
> >> now.
> >>
> >
> > The GB cannot make decisions based on the bylaws if we do not have
> > bylaws. Until then, it appears that the original interim rules
> > and conventions still apply. This is how some of
> > the prospective JDK8 projects (like lambda) have already been set up.
> > Oracle could insist on doing the same for jdk8 itself, despite the GB.
> > However, if jdk8 escapes the upcoming new bylaws, then primary jdk
> > development may continue to operate under the old interim conventions
> > for years.
> >
> > -Doug
> >
> 
> Can you explain what is wrong with the current conventions?  I can
> think of many things I believe could be improved with the current
> OpenJDK project, but the conventions for creating new projects doesn't
> factor high on the list.  I'm not saying the current rules are
> perfect, but they've worked for the last four years and allowed a
> number of projects to be started by non-Oracle contributors (the
> porting project, IcedTea, the Mac OS port).  I also hardly think you
> can claim that lambda is a OpenJDK8 project setup under the old
> guidelines as it was setup as an OpenJDK7 project, and is only now
> part of OpenJDK8 due to the decision to move it.  In short, continuing
> to operate under the interim conventions doesn't seem a problem to me.
> 
> I really don't see how it's better for the OpenJDK project for
> development to take place in private internal forks, jeopardising the
> future of the project as a whole, than to allow OpenJDK8 to 'escape
> the upcoming new bylaws'.  I've noted that you still do most of your
> development outside of the OpenJDK project, so maybe you aren't aware
> of the problems that we've experienced with getting Oracle developers
> to work in the open and the issues that have resulted from not having
> access to some of the early stages of OpenJDK7 development.  For some
> bug IDs, the corresponding changesets are simply unavailable as they
> pre-date the repositories.  It is at least understandable that the
> start of OpenJDK7 development is missing, due to the change from a
> proprietary development model, but I see absolutely no good reason to
> deliberately do the same with OpenJDK8.
> 
> I really don't see how this will force Oracle to do anything either.
> Most Oracle developers seem to prefer working in private repositories
> and on private mailing lists, presumably because it's the pre-existing
> dominant culture.  This is even noted in the minutes; 'Adam disagreed,
> noting that working behind closed doors is actually a more comfortable
> mode of operation for most Oracle engineers'.  This is going to hurt
> external engineers, like us at Red Hat instead.  Who knows when we
> will get access to OpenJDK8 development work and in what state it will
> be by then?  You're giving Oracle a "get out of jail free" card to do
> private development and use this board decision as an excuse.
> --
> Andrew :-)
> 
> Support Free Java!
> Contribute to GNU Classpath and IcedTea
> http://www.gnu.org/software/classpath
> http://icedtea.classpath.org
> 
> PGP Key: F5862A37 (https://keys.indymedia.org/)
> Fingerprint = EA30 D855 D50F 90CD F54D  0698 0713 C3ED F586 2A37