openjdk build pages

[Omair Majid]

* dalibor topic <dalibor.topic at oracle.com> [2014-05-22 13:13]:
> On 22.05.2014 01:44, Omair Majid wrote:
> >* dalibor topic <dalibor.topic at oracle.com> [2014-05-21 05:15]:
> >>and update releases that we can't work on as part of OpenJDK (like
> >>7u55),
> >
> >I am not sure I follow. If you can commit the source to the repository
> >and tag it, why can't you create a source bundle for those tags?
> 
> It doesn't make a lot of sense for a Project to provide source code archives
> for releases that are not releases actually produced by the Project.

Okay, can we please start doing those releases? I mean, I think everyone
agrees that people want to use the non-vulnerable versions of OpenJDK
and would much rather use, say, the 7u55 release rather than 7u40.

> >As for benefit, just today I saw people asking on #openjdk about where
> >to get source bundles.
> 
> They are linked off the front (and any other OpenJDK web site) page.

Didn't this thread start with the issue that there are no 8u bundles
published? If they are, please ignore me.

> >And they complained that using source control to
> >get a release bundles is too hard (and shouldn't be necessary).
> 
> I'm not sure why one would consider
> 
> hg clone http://hg.openjdk.java.net/jdk7u/jdk7u40; cd jdk7u40; bash
> get_source.sh
> 
> to be too hard compared to
> 
> wget http://www.java.net/download/openjdk/jdk7u40/promoted/b43/openjdk-7u40-fcs-src-b43-26_aug_2013.zip

Because the first one doesn't work with 8u. There is no way to get the
latest 8u20, for example.  Doing this in 8u will get you the latest 8u
in development. It wont get you the latest tag for 8u20 (or 8, for that
matter).

Also, get_source.sh checks out the latest sub-repos, not the one
matching the parent tag. With all due respect, if you can make this
mistake, how do you expect users to avoid it?

> - it seems to be fewer characters, even. ;)

But it's completely unusable for someone trying to get a known-good
what-was-realeased-as-8u20 1 year from today.

> >Also, if you think users have problems distinguishing 7u60 from 7u55,
> >can you imagine the problems they will have trying to find the
> >real/final tag for 7u55 in the repos?
> 
> I don't see that as problematic if it's documented.
>
> For example, the Project page for JDK 7 Updates could eventually say
> 'Hooray! 7u60 was released. The source code available in the jdk7u/jdk7u60
> repository with the  release being tagged with something-or-other. Use hg
> clone [snip] get_source.sh to check it out.'

Does this mean this will be fixed?

> > And how some tags do not exist in
> >some repos at some points in time? [2].
> 
> As soon as you have more then one forest covering the same Project, some
> tags won't exist in some repo at some point in time, while they exist in
> another, because commits to repositories are not atomic - not at the forest
> level, and not at the inter-forest-synchronization level.

Exactly. It's much easier to figure out what the latest release is if
you have a canonical source bundle available rather than hunting
through all the various project repositories trying to find out where a
tag is.

Thanks,
Omair

-- 
PGP Key: 66484681 (http://pgp.mit.edu/)
Fingerprint = F072 555B 0A17 3957 4E95  0056 F286 F14F 6648 4681