On Fri, Nov 23, 2018 at 11:41:10AM +0100, Mathieu Malaterre wrote: > As you've noticed the ABI breakage occur in between two minor uploads > -1 and -2. So I suspect this may confuse reader that bug be reported > against src:gdcm, since obviously not a single line change could have > occurred in between two minor uploads.
I don't understand this sentence of yours. The debian revision in a version doesn't convey any particular meaning about the "dimension" of the changes. In fact going from 2.8.7-1 to 2.8.7-2 could techinically ship a completely different upstream tarball with completely different code in it, still you would call that a "minor upload"? So everything could have happened in between. > Trying to read the changelog of 2.8.7-2 I only find reference to py3 > changes, however digging a bit in the d/control Deps I can see a > switch from vtk6 to vtk7: > > https://salsa.debian.org/med-team/gdcm/commit/616785342cdfc6747125a3f505af0b985d4d8fee > > Since libvtkgdcm2.8 is a c++ project, I suspect that any change of any > inherited c++ class will change the ABI. I would suggest that > libvtkgdcm2.8 be removed, and a new binary package libvtk7gdcm2.8 be > uploaded instead. Please also read the rest of my email, that's what I deduced and recomended as well (apart from suggesting a particular name for the binary). Also, before doing that somebody should check all the other binaries, and verify they didn't change their ABI either. BTW, it bothers me quite a lot that such a big change as moving from VTK6 to VTK7 has not been mentioned in the changelog at all. -- regards, Mattia Rizzolo GPG Key: 66AE 2B4A FCCF 3F52 DA18 4D18 4B04 3FCD B944 4540 .''`. more about me: https://mapreri.org : :' : Launchpad user: https://launchpad.net/~mapreri `. `'` Debian QA page: https://qa.debian.org/developer.php?login=mattia `-
signature.asc
Description: PGP signature
