Hi,
Well I addressed this in the Pipeline build. This is one job and it fails if
any of the builds in the pipeline fail. If you committed something and it
fails, the chances are high, that it’s related to your changes. I usually
double check the build after committing things.
But I guess the biggest problem is that people don’t seem to be looking at the
output the build produces. Eventually even have filters on the commit list to
filter out failing build emails. AND I think it’s a real problem that I hear a
lot of “But it works on Ant” as an excuse. If I turn off tests, build without
distribution and examples, I will also get a successful build with Maven, but I
still won’t get running applications (This is what the integration-test module
tests).
I guess implementing such an auto-branch-and-revert will not be easy … so I am
thinking of just doing it manually. I am currently extremely worried about the
significant drop in email and commit activity since the breaking changes (at
least that’s my impression). Right now, someone working with the SNAPSHOT
version will not be able to get a working application and that has been that
way for about two weeks. That’s inacceptable from my point of view (
Chris
Am 05.04.17, 13:45 schrieb "yishayw" <yishayj...@hotmail.com>:
Probably a good idea. I have two concerns. One is how to make it clear to
the
committer that s/he needs to make things work. When the build is broken it's
at least clear that it needs to be fixed ASAP. The other is how to identify
the faulty component. If Falcon is broken problems might manifest in asjs.
--
View this message in context:
http://apache-flex-development.2333347.n4.nabble.com/FlexJS-Discuss-Automate-stabilizing-develop-tp61014p61015.html
Sent from the Apache Flex Development mailing list archive at Nabble.com.
