The relevant PMC isn't the Chukwa PMC, it's the Hadoop PMC. And I'm not comfortable doing a release without their thumbs-up. You're welcome to publicize the candidate build if you like -- the .tgz is available online.
But I'm not sure I understand the rationale you outline. Why would people want to try 0.2 rather than trunk, or waiting for 0.3? Realistically, I don't think 0.2 is more stable than trunk at the moment. And the documentation reflects 0.3-trunk, not 0.2. --Ari On Tue, Jul 28, 2009 at 7:36 AM, Jerome Boulon<[email protected]> wrote: > I would still release 0.2 just just in case someone want to try Chukwa. > With 3 +1, no -1 and since there's no PMC for Chukwa I guess we can release. > /Jerome. > > > On 7/28/09 2:17 AM, "Ariel Rabkin" <[email protected]> wrote: > > Hi all. > > We never did get PMC signoff for the 0.2 release, and I think we've > done enough since then to be worth trying to release 0.3, instead. > It's in use at Berkeley, and people are pretty happy. Plus there's a > fairly large chunk of code for SALSA that ought to be released. > > I think there's just a handful of blockers before release. > > - We need tests for CHUKWA-344. CHUKWA-348 might also be a blocker. > > - I want to revise and reorganize the admin guide. In particular, I > want to pull out the hicc stuff into a separate document, and also > pull out some other sections into a "quickstart" and/or "overview" > document. > > - Some refactorings and such that we might as well do before > branching. (notably CHUKWA-355 and CHUKWA-345) > > My hope would be to feature freeze+branch in early August, and roll a > release shortly thereafter. Thoughts? > > --Ari > > -- > Ari Rabkin [email protected] > UC Berkeley Computer Science Department > > -- Ari Rabkin [email protected] UC Berkeley Computer Science Department
