Joerg Jaspert <[email protected]> writes: > would be nice if Lintian could have an E: level for an easy to detect > FTBFS. If there is a CMakeCache.txt file in the .diff.gz and as such in > the build tree, cmake will simply refuse to build the package and > outright fail: > > CMake Error: The current CMakeCache.txt directory > /tmp/buildd/mysql-connector-cpp-1.0.2~20081215/CMakeCache.txt is > different than the directory > /home/rene/mysql-connector-cpp-1.0.2~20081215 where CMackeCache.txt was > created. This may result in binaries being created in the wrong > place. If you are not sure, reedit the CMakeCache.txt > make: *** [configure-stamp] Error 1 > > To not have a chance of too many false positives, I would suggest > looking directly into the diff, so avoid upstream source. And do not > count the error as soon as you find that filename anywhere in the rules > file (that assumes the Maintainer takes care of it somehow).
This is fairly trivial to do if we don't look in debian/rules and a bit harder to do if we do. Before I write the code to scan debian/rules, though, I wanted to check if you think there's some legitimate case where someone would create or modify this file in the diff and then remove it in debian/rules. I can't think of one, and hence am leaning towards just always flagging it as an error if that file is present in the diff. -- Russ Allbery ([email protected]) <http://www.eyrie.org/~eagle/> -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected]
