On Friday, August 14, 2015 08:37:31 AM Scott Howard wrote: > On Fri, Aug 14, 2015 at 4:05 AM, Simon McVittie <s...@debian.org> wrote: > > On Tue, 11 Aug 2015 at 18:47:58 -0400, Scott Howard wrote: > >> Package renamed to libmuparser2v5. > >> See patch: > >> http://anonscm.debian.org/cgit/debian-science/packages/muparser.git/patch > >> /?id=5fb47ad4af6a7e4cdddbb078b7159d552c0e1f86> > > This is unusual: you've changed the SONAME to libmuparser.so.2v5. > > For other packages in this transition, the approach that has been taken > > (as recommended in the mass bug filing) was to rename the *package* to > > libmuparser.so.2v5, but leave the SONAME at libmuparser.so.2. > > > > For instance, here's a similar change to liborigin2: > > <http://launchpadlibrarian.net/213540578/liborigin2_2%3A20110117-1build4_2 > > %3A20110117-1ubuntu1.diff.gz> (Note that the lintian override in that > > patch is unnecessary in Debian; lintian has been changed to not complain > > about the v5 suffix.) > > > > Release team: is it a problem that the transition has been done > > differently > > in this way? > > Thank you - this is my mistake, I modelled it after this NMU: > https://bugs.debian.org/cgi-bin/bugreport.cgi?att=1;filename=libbitcoin.nmu. > debdiff;bug=791092;msg=19 Which I originally thought bumped the SONAME as > well. We can restore it before upload to unstable. > ~Scott
Since this has no C++ build-deps is can go ahead to unstable without RT ack. Scott K
