So there has been discussions off-and-on about assembling a dedicated QA team for the Haiku project that could lend a hand in testing Haiku release candidates, etc… that would ensure improved quality of Haiku release to release. PulkoMandy has a desire for a QA team to be up-n-running and ready to tackle testing of release candidate(s) for Haiku B2 when its planned for release sometime next year. So this post is meant to have an active conversation about how such a QA team/process would work.
As you know, the Haiku project is very agile and doesn’t currently have a formal development process per say as the contributors work on what they feel is important and is inline with their interests and confidence level. There is a monthly report that goes out and I guess one could look at each month’s worth of development as a single sprint and maybe QA could just sample a nightly build at the end of each month and peform regression testing and report back to the Haiku development team.
Currently there are some manual test cases documented and archived in various places and would need to be collected and cataloged in a single place. I was thinking that it would be great to get an instance of TestLink hosted within Haiku’s ecosystem and migrate the (whatever existing) documented test cases and organize them in a way that makes sense for ease of maintenance, execution, etc…
So who else out there within the Haiku community has a QA background and/or is interested in such an endeavour?
Just an FYI… there is a haiku-qa IRC channel (I hangout there alone now) as well as a mailing list (https://www.freelists.org/archive/haiku-qa/).