On 2013-04-25 22:09, Jakub Wilk wrote: > * Andreas Beckmann <a...@debian.org>, 2013-04-25, 21:27: >> trying to overwrite shared '/usr/lib/debug/usr/lib/libffi.so.5.0.10', >> which is different from other instances of package libffi5-dbg:i386 > > #650106 > >> Maybe this shouldn't have been MA:same. > > There's no problem with -dbg packages being MA:same. They just need to > use build-id-based paths. dh_strip does it for you in compat >= 9.
So "maybe libffi5-dbg shouldn't have been M-A:same for wheezy since it was too late to bump debhelper compat to 9". Should we try to fix this for wheezy (by temporarily dropping MA:same on the -dbg package)? Fix can go via unstable. >> trying to overwrite shared >> '/usr/share/doc/libdmtx-dev/examples/net/Libdmtx.Net.Test/DmtxTest.cs.gz', >> which is different from other instances of package libdmtx-dev:i386 > > I can't reproduce this in unstable. it's a gzip problem ... the uncompressed files are identical 5211e14d78f209547869f5f0511e8962 DmtxTest.cs.gz:i386 50d26e0c550cd9c99223abf2dbe7b408 DmtxTest.cs.gz:amd64 107f0f804c01fc96697554b44403da17 DmtxTest.cs:amd64 107f0f804c01fc96697554b44403da17 DmtxTest.cs:i386 If I rebuild the wheezy package in sid for amd64, I get the following md5sum: 5211e14d78f209547869f5f0511e8962, so maybe there was something badly configured or outdated on the buildd? https://buildd.debian.org/status/logs.php?pkg=libdmtx&arch=amd64 0.7.2-2+b1 (sid) Maybe-Successful 2012-04-03 18:36:35 brahms 6m 9.36 MB Anyway, libdmtx was binNMUed, so this "problem" might clear automatically for a no-change rebuild. Jakub, thanks for filing / looking up the other bugs. Maybe some of them are affected by some gzip problem, too, as the problematic files were compressed ones. Andreas -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/517b2943.7020...@debian.org
