OpenJDK (Java) migrating from Mercurial?

[Scott Palmer]

> On Nov 15, 2019, at 6:19 PM, František Kučera <konference at frantovo.cz> wrote:
> 
> Dne 26. 07. 19 v 17:19 duvall at comfychair.org <mailto:duvall at comfychair.org> napsal(a):
>> On Fri, Jul 26, 2019 at 03:03:12PM +0200, Arne Babenhauserheide wrote:
>> 
>>> Marek Lukáš via Mercurial <mercurial at mercurial-scm.org <mailto:mercurial at mercurial-scm.org>> writes:
>>> …
>>> From this: https://mail.openjdk.java.net/pipermail/discuss/2019-July/005096.html <https://mail.openjdk.java.net/pipermail/discuss/2019-July/005096.html>
>>> …
>>> It looks like most of the arguments for the move are not based in
>>> delivering improvements with the minimal amount of breakage, but rather
>>> in moving to git at any cost.
>> It seems likely that's the case: some unrelated team at Oracle has made the
>> decision that everything will be git, and now engineering teams (willing or not)
>> have to implement that decision, regardless of external consumers.  With Oracle,
>> if you're not a paying customer, you're not valued very much, if at all.  And
>> I'm sure that Oracle (again, as a company, not the individual teams and people
>> who interact with external communities) feels that they own OpenJDK as much as
>> they own Java, and that the community participation in governance is simply an
>> impediment to them doing whatever they feel they need to do.
>> 
>> There was strong pressure when I was still there to move the Solaris repos to
>> git.  In fact, the default stance (dictated by some small team way up in the
>> CTO's office, I think) was git for any external participation.  You were
>> expected to have a repo on github, though they relented and allowed us to
>> set up a mirror of the internal mercurial repo (for our opensource bits).  At
>> least one other repo gave up and converted.  I wouldn't be surprised if more
>> did.
>> 
>> Danek
> BTW: there is a new discussion „New candidate JEP: 369: Migrate to
> GitHub“ in the OpenJDK mailing list:
> <https://mail.openjdk.java.net/pipermail/discuss/2019-November/thread.html <https://mail.openjdk.java.net/pipermail/discuss/2019-November/thread.html>>
> 
> Some comments:
> 
>> I think we really need to get back to this proposal in an year or two
>> when we understand more of the implication
> 
> <https://mail.openjdk.java.net/pipermail/discuss/2019-November/005200.html <https://mail.openjdk.java.net/pipermail/discuss/2019-November/005200.html>>
> 
>> It worries me a little that an entire DVCS swap is being proposed
>> without, apparently, any attempts to speed up Mercurial operations
>> having being tried
> 
> <https://mail.openjdk.java.net/pipermail/discuss/2019-November/005195.html <https://mail.openjdk.java.net/pipermail/discuss/2019-November/005195.html>>
> 
> Franta

Yes, but I expect it is a done deal regardless.  The discussion smells like it’s just there to show they asked for input, but the decision was finished long ago.  OpenJFX already moved to Git. ( I had to deal with Git insanity when I submitted patches.  Such an awkward system to work with when you are used to Mercurial. It was painful.)
They do have a lot of automation based on GitHub though.  I’ll give them points for that.  

Scott


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mercurial-scm.org/pipermail/mercurial/attachments/20191115/8df63e16/attachment-0002.html>