Hi, Am Freitag, den 28.08.2015, 15:44 +0200 schrieb Hector Oron: > Not sure if sufficient, but at least it is interesting test case to > check. I would also like PPA-like (bikesheds) scenario to be tested, > even the case of multiple chained PPA-likes, but I do not think we > currently have real data. Also considering Haskell packages as > initial > set sounds good to me, as those have quite tight versioned > dependencies having extremely long dependency resolution chain which > exposes aptitude solver bug. Can’t we got at it more pragmatically? If wanna-build determines that the build dependencies are satisfiable, then they are. It does, however, not matter with what set of packages wanna-build (or rather, dose) came to that conclusion. On the buildd side, if apt/aptitude does not find a solution, it is already wrong. So if, as done in Johannes’ patch, in this case sbuild tries again using a different solver, I cannot see how that could make matters worse in any way than they are now. So in the sense of „what could possibly go wrong“, I don’t see what you expect from Johannes and why we cannot simply make the buildds try the more advanced solver, and replace a few failed builds by succeeding builds. If there are obscure cases where the advanced solver is indeed invoked, and where it pulls in too many packages, and where that actually makes a difference to the build, they can be fixed as we go. And it sounds to me that such a problem would be a bug in the package (missing Build -Conflicts etc.) Don’t let perfect be the enemy of good! After all, this is a real problem (I get hourly mails about failed builds of various packages) and it seems there is a fix... (Or maybe I’m misunderstanding you and the issues you are discussing.) Greetings, Joachim -- Joachim "nomeata" Breitner Debian Developer nomeata@debian.org | ICQ# 74513189 | GPG-Keyid: F0FBF51F JID: nomeata@joachim-breitner.de | http://people.debian.org/~nomeata
Attachment:
signature.asc
Description: This is a digitally signed message part