Merge lp:~jtv/launchpad/translationtemplatesbuild into lp:launchpad/db-devel
Status: | Merged |
---|---|
Approved by: | Jeroen T. Vermeulen |
Approved revision: | no longer in the source branch. |
Merged at revision: | 9793 |
Proposed branch: | lp:~jtv/launchpad/translationtemplatesbuild |
Merge into: | lp:launchpad/db-devel |
Diff against target: |
423 lines (+242/-23) 12 files modified
database/schema/comments.sql (+5/-0) database/schema/patch-2208-99-0.sql (+17/-0) database/schema/security.cfg (+2/-0) lib/lp/buildmaster/model/buildfarmjob.py (+1/-1) lib/lp/code/doc/branch.txt (+1/-0) lib/lp/code/model/branch.py (+9/-7) lib/lp/translations/configure.zcml (+11/-0) lib/lp/translations/interfaces/translationtemplatesbuild.py (+39/-0) lib/lp/translations/model/translationtemplatesbuild.py (+71/-0) lib/lp/translations/model/translationtemplatesbuildbehavior.py (+4/-2) lib/lp/translations/model/translationtemplatesbuildjob.py (+16/-13) lib/lp/translations/tests/test_translationtemplatesbuild.py (+66/-0) |
To merge this branch: | bzr merge lp:~jtv/launchpad/translationtemplatesbuild |
Related bugs: |
Reviewer | Review Type | Date Requested | Status |
---|---|---|---|
Stuart Bishop (community) | db | Approve | |
Michael Nelson (community) | code | Approve | |
Robert Collins | db | Pending | |
Review via email: mp+34952@code.launchpad.net |
Commit message
TranslationTemp
Description of the change
= TranslationTemp
Buildfarm work. This creates a new TranslationTemp
http://
Until we actually move over to the new data model, the TranslationTemp
The big picture I may still be missing is who's responsible for creating what objects. Implementing makeJob for the old-style setup triggered some upheaval in the TranslationTemp
Jeroen
We did this review from a paste while LP was down. In summary, r=me assuming you'll not land this until you've got a second branch that ensures the builder (current builds/history) pages display these correctly (ie. you'll need to ensure BFJ.getSpecific Job() finds the appropriate adapted.
Thanks!
10:09 < jtv> noodles775: thanks for the review btw… I _would_ point you to my buildfarm work now, but Maintenance. : /code.edge. launchpad. net/~jtv/ launchpad/ translationtemp latesbuild/ +merge paste.ubuntu. com/490815/ lateBuild via ZCML (rather t platesBuild inherit from IBuildFarmJob - like IPackageBuild does)? platesBuild also an IBuildFarmJob? I'm not fussed either, I was just surprised. Job() works for TranslationTemp latesBuild. Job() works for the translation template builds)
/
10:10 * noodles775 breathes a sigh of relief ;)
10:11 < noodles775> jtv: if you can paste your MP (and a diff), I can do that too :)
10:13 < jtv> noodles775: the MP is at https:/
/34952
10:13 < noodles775> jtv: sorry, I meant paste the text of your MP to paste.....com :)
10:13 < jtv> noodles775: yes, I'm otp so a little slow sorry :)
10:22 < jtv> noodles775: my diff is here: http://
10:31 < noodles775> jtv: Was that a decision to allow both interfaces for TranslationTemp
han having ITranslationTem
10:32 < jtv> noodles775: barely registered. :) I tend to avoid interface hierarchies because they can get a little confusing in combination with class hierarchies, but if that's wrong I'll happily change it.
10:32 < jtv> (not that zcml can't get a little confusing…)
10:33 < noodles775> I'm not sure if its better or worse... I think of it with the question: Is the interface ITranslationTem
10:34 < jtv> We're sort of probing the boundaries between is-a and has-a here, aren't we?
10:35 < jtv> I guess in python we're doing inheritance, just not in the db—in which case perhaps I should change this after all.
11:09 < noodles775> jtv: sheesh, it'll be nice to get rid of the BFJOld stuff (and makeJob). A question, in your makeJob() method, you don't add the TTBJ to the store... am I missing something? I'm wondering if you meant to call TTBJ.create() there?
11:10 < jtv> noodles775: ah good point, thanks
11:11 < jtv> noodles775: ah! Actually I do add it to the store.
11:11 < jtv> noodles775: in terms of db storage, the BranchJob _is_ the TTBJ
11:31 < noodles775> jtv: ah, so TTBJ isn't a storm class itself. Right.
11:31 < jtv> just one more reason to clean out these stables :)
11:41 < noodles775> jtv: so r=me, assuming that you won't be landing this without further branches (ie. to ensure BFJ.getSpecific
11:41 < noodles775> Thanks!
11:42 < jtv> noodles775: that's actually a lot more than I was hoping for, thanks! Any other points I need to cover?
11:43 < noodles775> jtv: er, you could refactor the code and get rid of IBuildFarmJobOld? ;)
11:44 < noodles775> I can't think of any other things... the above will make it self apparent when you look at current builders or builder histories after having dispatched TTBs.
11:45 < noodles775> (sorry, where "the above" is ensuring BFJ.getSpecific
11:45 < jtv> OK, so I'll move on to that next then.