[11u]: 11.0.4 release schedule
Lindenmaier, Goetz
goetz.lindenmaier at sap.com
Wed Apr 24 10:55:08 UTC 2019
> Am 24.04.2019 um 12:03 schrieb Aleksey Shipilev <shade at redhat.com>:
>
>> On 4/24/19 11:47 AM, Langer, Christoph wrote:
>> I've now published the following schedule for 11.0.4 on the Wiki [0]:
>>
>> March 2019: jdk11u-dev repo open (tag: jdk-11.0.4+0)
>> Tuesday, April 30 2019: Branch jdk11u-dev to jdk11u
>> Wednesday, May 1 2019: First build (tag: jdk-11.0.4+1)
>> Wednesday, May 29 2019: RDP2
>> Wednesday June 26 2019: Last tag before code freeze
>> Tuesday, July 16 2019: GA (tag: jdk-11.0.4-ga)
>>
>> Is everyone in agreement with that?
>>
>> Note: It means that we'll branch next week from jdk11u-dev to jdk11u!
>
> That also means outstanding 11.0.4-oracle things are better to get to 11u-dev within a week, right?
> https://bugs.openjdk.java.net/issues/?filter=36409
It‘s good if they make -dev, but they can go to jdk11u if we miss the branch.
>
> I have my doubts about having 2.5 months gap between fork and release. That feels too early. Or
> maybe not, given there is a pending 13 freeze in between, and also summer vacations.
Actually for the community it’s only eight weeks. After that the repo is frozen.
I think this time is needed to bring in the last non-critical changes, build, test, fix bugs,
push the bug fixes and re-iterate.
>
>> Furthermore, I'm questioning myself, whether we want to explicitly have an RDP2 date? Pushes that target the jdk11u repo need to use the "critical" labeling and should meet the stricter rules for release stabilization anyway (see section "Fix Approvals" in the link below). Any opinions on that one?
>
> IMO, if we are not pulling jdk11u-dev -> jdk11u automatically and regularly, then there is little
> sense in having the formal RDP2.
We only pull in that direction once per release.
The other way is pulled weekly.
> Maintainers would have to judge if change is risky for
> stabilization forest, every step of the way anyway.
This should be expressed by the priority of the bug. Naturally, when allowing a change to jdk11u, Andrew will double check, I assume.
Best regards,
Götz
>
> -Aleksey
>
More information about the jdk-updates-dev
mailing list