On Sun, 2015-02-22 at 22:09 -0800, Bill Deegan wrote: > Greetings, > > With help from Dirk and Russel we're green (no failures in runtest.py) on > all buildbot slaves. > > I'd like to suggest that going forward any failures after a patch must be > fixed before any other patch is applied. > > Any objections? >
That strikes me as an excellent management approach. Of course the downside is that pull requests can only be applied as fast as "all green" appears on the Buildbot console after a merge into mainline. Fine if no errors, but potentially problematic in the face of awkward new reds. It would seem to me that there actually needs to be two CI instances: a testbed instance which operates on a newly created repository that is transient. A pull request is tried on this ephemeral repository, if green then a merge of the pull request into mainline is allowed. If red then preserve the CI results for analysis, but do not merge into mainline. This way only passing pull requests are ever merged into mainline and there is never a hold up. -- Russel. ============================================================================= Dr Russel Winder t: +44 20 7585 2200 voip: sip:[email protected] 41 Buckmaster Road m: +44 7770 465 077 xmpp: [email protected] London SW11 1EN, UK w: www.russel.org.uk skype: russel_winder
signature.asc
Description: This is a digitally signed message part
_______________________________________________ Scons-dev mailing list [email protected] https://pairlist2.pair.net/mailman/listinfo/scons-dev
