On Thu, Oct 08, 2015 at 04:55:35PM +0200, Holger Levsen wrote: > Hi, > > On Donnerstag, 8. Oktober 2015, Santiago Vila wrote: > > I've seen several cases where a package is considered not reproducible > > just because the build environment changed between build1 and build2. > > this should not happen… end of the story. If it happens, its a bug in the CI. > > and today it also happened because of adding the new builders... (i'll spare > myself explaining the exact details, just we dont add builders every day, so > meh.) > > > However, I can think of some workarounds: > > yes, we too… > > I'm sorry but I'm severely overloaded and these problems are well known, well > discussed and being worked on. Explaining them here again, just takes away > time to work on these issues.
Well, long story short: the build script compares the two .buildinfo just after the build and if Build-Environment differes it re-runs the build a 3rd time. If that fields differs again (this might be the case for very long running builds that spawn several dinstall and different toolchain packages change during the day) it just gives up and mark the package as unreproducible > It might be helpful to join irc to discuss with us in realtime. Agree, be assured discussing some stuff on IRC is really quicker than email. Though when it comes to #debian-reproducible following everything might be sometimes annoying due to the volume of the chats, but imho is still worthwhile :) -- regards, Mattia Rizzolo GPG Key: 66AE 2B4A FCCF 3F52 DA18 4D18 4B04 3FCD B944 4540 .''`. more about me: http://mapreri.org : :' : Launchpad user: https://launchpad.net/~mapreri `. `'` Debian QA page: https://qa.debian.org/developer.php?login=mattia `-
Description: PGP signature
_______________________________________________ Reproducible-builds mailing list Reproducibleemail@example.com http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds