8u/11u repo access and Jira changes

Aleksey Shipilev shade at redhat.com
Tue Feb 26 20:23:04 UTC 2019

On 2/26/19 9:04 PM, Andrew Haley wrote:
> On 2/26/19 7:46 PM, Aleksey Shipilev wrote:
> Maybe it's a cultural thing. I've worked on other large-scale free
> software projects (GCC, in particular) where it just wasn't an issue
> because people knew when they should be working on a release branch
> and it simply wasn't an issue, but equally it was obvious which were
> the release branches. Perhaps people working on OpenJDK aren't up to
> it. :-)

It partially is a cultural thing. I grew up in the culture that assigned all the responsibility to
the individual workers, and no fault was ever systemic; I saw how devastating that is for both the
workers and the system. I understand the reverse side of this: when everything is system fault,
there is no personal responsibility; equally devastating.

So, the limited set of power users that can push to the stable tree is a happy amalgamation of both
approaches: stable tree is "free for all, but not for everybody". ;)

Speaking of large software projects, I think Linux gets it right: maintainers "pull" changes into
upper repositories rather than technically allowing anyone from downstream to push on their own. For
all the perfect people who are working on Linux, that is a saner approach for a responsible
maintainer, in my mind. I think this basically happened with jdk8u/jdk8u and jdk8u/jdk8u-dev split,
and once again I question why do we need to deviate...


More information about the jdk8u-dev mailing list