The work in bug 523946 is against the scheduler db. Should this be marked as a duplicate or a dependent bug?

(In reply to comment #1) > The work in bug 523946 is against the scheduler db. Should this be marked as a > duplicate or a dependent bug? no, this bug is going to be for getting existing mobile code running in a buildbot-0.8.0 friendly way using schedulerdb

this is running at http://mobile-sandbox.build.mozilla.org:8010/builders/Maemo%205%20GTK%20mozilla-central%20build/builds/1 That is a scheduler master and build master with ACTIVE_BRANCHES set to []

Not sure how(more where) to stage this Notes: -Not too sure how to set up an automatic symlink from mobile_config_[production|staging].py -> mobile_config.py for the master setup script (haven't looked) -For some reason the HgPoller for mobile-browser was giving me trouble. I have disabled it but this is a definite blocker changes to the mobile_config.py files were done by running: cd buildbot-configs/mozilla cp ../mozilla2-staging/mobile_config.py mobile_config_staging.py cp ../mozilla2/mobile_config.py mobile_config_production.py sed -i -e "s/MOBILE_BRANCHES\['mobile-addonsmgr'\]/MOBILE_BRANCHES\['addonsmgr'\]/ ; s/MOBILE_BRANCHES\['mobile-tracemonkey'\]/MOBILE_BRANCHES\['tracemonkey'\]/ ; s/MOBILE_BRANCHES\['mobile-electrolysis'\]/MOBILE_BRANCHES\['electrolysis'\]/ ; s/MOBILE_BRANCHES\['mobile-trunk'\]/MOBILE_BRANCHES\['mozilla-central'\]/ ; s/MOBILE_BRANCHES\['mobile-1.9.2'\]/MOBILE_BRANCHES\['mozilla-1.9.2'\]/ ; s/'mobile-trunk'\:/'mozilla-central':/ ; s/'mobile-1.9.2'\:/'mozilla-1.9.2':/ ; s/'mobile-addonsmgr'\:/'addonsmgr':/ ; s/'mobile-tracemonkey'\:/'tracemonkey':/ ; s/'mobile-electrolysis'\:/'electrolysis':/ ; " mobile_config_staging.py mobile_config_production.py

Ran this overnight in staging: -linux worked 19 out of 23 times: -All failures are on QT during compile, buildbot acted as expected -did multiple Maemo 4, Maemo 5 GTK and Desktop builds -did compile Maemo 5 QT successfully once -win32 worked 2 out of 3 times -the failure was a failure during compile, buildbot acted as expected -macosx worked 2 out of 3 times -the failure was a compile failure, buildbot acted as expected. It looks like the windows and osx issues were a code issue with 'mozilla-central/content/html/content/src/nsHTMLMediaElement.cpp'. I have verified that the nightly does build on these platforms from a forced build earlier yesterday I am going to fix the change source then ask for review on this patch. This patch will land for staging in entirety, but will land as a subset for production as we can't disable MultiL10N or regular L10N repacks.

(In reply to comment #2) > (In reply to comment #1) > > The work in bug 523946 is against the scheduler db. Should this be marked as a > > duplicate or a dependent bug? > > no, this bug is going to be for getting existing mobile code running in a > buildbot-0.8.0 friendly way using schedulerdb actually, the change source being missing wasn't as easy to fix as I had thought. I decided to move my efforts to bug 523946 as it will take approximately the same amount of time as the efforts of this bug. Patch being uploaded to that bug shortly