www-jcp-open mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Henri Yandell" <bay...@apache.org>
Subject Re: stubbed out implementations of jcp-managed interfaces
Date Fri, 26 Oct 2007 18:08:20 GMT
On 10/26/07, Rick Hillegas <Richard.Hillegas@sun.com> wrote:
> The following issue has come up on the Derby developer list and I would
> like your advice about how to proceed:
> In order to build a releasable distribution of Derby, a developer needs
> to compile Derby's small device support. Right now, this involves
> finding an implementation of the CDC foundation classes and the JSR 169
> database interface classes--both of which implementations are licensed
> under JCP rules. As far as I know, the only implementations of these
> interfaces are available from companies which require that you accept
> their additional, encumbering licenses. Rather than sending every
> release manager out on this hunt, I'd like to be able to check in enough
> Apache-compatible machinery that the Derby build would always
> automatically compile the small device support.
> One possibility would be to create our own stubbed out implementation of
> the CDC foundation and JSR 169 classes. A warning, however, has been
> raised. We are worried that the JCP rules may not allow us to create a
> stubbed out implementation which is only useful for compilation
> purposes. Is that also the opinion of the jcp-open@apache.org list?
> I would appreciate your advice on how I can get my hands on unencumbered
> compile-time machinery so that I can simplify the Derby build.

I thought Geronimo had done this a lot - you can use the spec to build
it but you can't use the javadoc as that is copyrightable. As long as
we don't release it claiming it is a certified implementation of the
JSR, then I think we're fine.

Will let others chime in if they think different.

We've long talked about having a central location for all of these
jars to sit so we're not all playing hunt-the-jar when its time to
find such a stub.


View raw message