7u Source Bundle

[Dalibor Topic]

On 5/8/13 9:14 PM, Kurt Miller wrote:
> On May 8, 2013, at 7:07 AM, Andrew Hughes <gnu.andrew at redhat.com> wrote:
>> ----- Original Message -----
>>> On 4/30/13 6:50 PM, Kurt Miller wrote:
>>>> The 7u source bundle [1] is quite outdated now (7u6 Aug 2012). Would it be
>>>> possible to get it updated to 7u21? This bundle is helpful to bsd-port
>>>> where we download the 7u source bundle and then apply our changes from
>>>> bsd-port to this.
>>>
>>> I'd expect that an updated bundle would be published for 7u40. I think that
>>> would be a better starting point, if we started doing that, then starting
>>> with 7u6 now and cumulatively applying past patches.
> 
> I'm confused. I figured it would just take pulling jdk7u or jdk7u-dev by a tag and creating a tarball of the result and updating the source bundle page. Am I missing something?

Yes, the (lack of a) release process for releases that were not developed within this Project. 

That may sound to abstract, so let me try to make it less so:

When OpenJDK 7u makes a release, it goes through a long drawn sequence of steps, which
serve to deliver a solid release that can then be used as the basis for downstream builds.

Just pulling a tag from repo and putting it online does not necessarily deliver that. It may not be
buildable on some downstream platform or other, or have other issues that can only be detected 
through testing once changes have been integrated into mainline forest.

That means the release one ends up producing for OpenJDK 7u would lag behind the initial bulk integration 
by some measure because of the need for testing and potentially additional iterations of fixes. Let's assume 
it's weeks for the sake of discussion.

This is not a purely hypothetical discussion - OpenJDK 6 release model was actually very close to that. See 
http://hg.openjdk.java.net/jdk6/jdk6/raw-file/tip/ReleaseProcess.html for details.

So, the two choices seem to be to either throw an untested and potentially unbuildable source code over the wall,
or to spend a fair amount of extra time stabilizing a release that most downstream projects won't need by the point 
it's available because they will have adapted and integrated the changes made available on jdk7u-dev in their own
releases already. None of the choices seems all that great.

> I like the idea of *BSD support in IcedTea. Now that Mac OS X support is in jdk7u, the number of pure BSD patches is more manageable. If bsd-port were merged into jdk7u or 8 first, I think basing our packages off IcedTea would be a fine idea. Keeping *BSD support out of the main line is not sustainable, in my opinion.

I think you should follow the example of the PPC/AIX port Project and look at their JEP for inspiration.

cheers,
dalibor topic

-- 
Oracle <http://www.oracle.com>
Dalibor Topic | Principal Product Manager
Phone: +494089091214 <tel:+494089091214> | Mobile: +491737185961 <tel:+491737185961>
Oracle Java Platform Group

ORACLE Deutschland B.V. & Co. KG | Kühnehöfe 5 | 22761 Hamburg

ORACLE Deutschland B.V. & Co. KG
Hauptverwaltung: Riesstr. 25, D-80992 München
Registergericht: Amtsgericht München, HRA 95603
Geschäftsführer: Jürgen Kunz

Komplementärin: ORACLE Deutschland Verwaltung B.V.
Hertogswetering 163/167, 3543 AS Utrecht, Niederlande
Handelsregister der Handelskammer Midden-Niederlande, Nr. 30143697
Geschäftsführer: Alexander van der Ven, Astrid Kepper, Val Maher

Green Oracle <http://www.oracle.com/commitment> Oracle is committed to developing practices and products that help protect the environment