Regarding replacing tinderbox, I think there aren’t any technical reasons that can’t be solved.

I don’t yet know how to do log parsers, and of course, logs need to be compressed on disk, and likely written outside the master dirs. I think that that should be subject to my next git branch.

I do have ideas on how to do tree rule messages and trees in general, though I didn’t start with that yet.

The question goes further though. Having the webdev piece on top of django is likely going to put us into a better place than tinderbox. It doesn’t put us into a place where the webdev piece is a real open source project with outside contributors, though. Maybe we have to bite that bullet to just get our development infrastructure improved and fix recent regressions like the blame column. But I anticipate a wider and more rational discussion on that.

Right now, this is my personal baby, and I try to make it look like an attractive replacement for tinderbox, but whether it is or not is up to discussion.

I started a similar project “Millicent” (which we discussed in the infrastructure improvements subject in m.d.planning) but I think I like your approach better.

The main thing I was concerned with was making a drop-in replacement for Tinderbox, which maybe would be better to do as a plug-in for Buildbot anyway and not have to create a third system.