On Tue, 30 Apr 2013 13:27:06 +0200, Diego Biurrun <[email protected]> wrote: > On Tue, Apr 30, 2013 at 11:22:49AM +0200, Anton Khirnov wrote: > > On Fri, 19 Apr 2013 21:23:26 +0200, Diego Biurrun <[email protected]> wrote: > > > On Thu, Apr 11, 2013 at 09:10:02PM +0200, Anton Khirnov wrote: > > > > --- > > > > Changelog | 1 + > > > > doc/filters.texi | 125 +++++++++++ > > > > libavfilter/Makefile | 2 + > > > > libavfilter/allfilters.c | 2 + > > > > libavfilter/trim.c | 407 > > > > ++++++++++++++++++++++++++++++++++ > > > > tests/fate/filter.mak | 30 +++ > > > > tests/ref/fate/filter-atrim-duration | 2 + > > > > tests/ref/fate/filter-atrim-mixed | 5 + > > > > tests/ref/fate/filter-atrim-samples | 2 + > > > > tests/ref/fate/filter-atrim-time | 6 + > > > > tests/ref/fate/filter-trim-duration | 2 + > > > > tests/ref/fate/filter-trim-frame | 8 + > > > > tests/ref/fate/filter-trim-mixed | 10 + > > > > tests/ref/fate/filter-trim-time | 3 + > > > > 14 files changed, 605 insertions(+) > > > > create mode 100644 libavfilter/trim.c > > > > create mode 100644 tests/ref/fate/filter-atrim-duration > > > > create mode 100644 tests/ref/fate/filter-atrim-mixed > > > > create mode 100644 tests/ref/fate/filter-atrim-samples > > > > create mode 100644 tests/ref/fate/filter-atrim-time > > > > create mode 100644 tests/ref/fate/filter-trim-duration > > > > create mode 100644 tests/ref/fate/filter-trim-frame > > > > create mode 100644 tests/ref/fate/filter-trim-mixed > > > > create mode 100644 tests/ref/fate/filter-trim-time > > > > > > version bump missing > > > > It's bumped in a previous commit which adds the asetpts filter. > > > > Since they will be pushed in one batch, I don't see a need to bump the > > version > > twice. > > Then how can you tell the point at which filter X was added to the > library? I believe we have had this discussion before and library > users care about having such changes flagged. >
I don't recall any discussion about this specifically. On the contrary, it's been common practice for as long as I can remember that multiple bump-worthy changes pushed at once get only one version bump in total. Until now I've never seen anyone complain about it. Since the git master was never at the state where it had one filter but not the other, I fail to see any reason to have two bumps. (and I must add that the whole practice of bumping version when adding new codecs/(de)muxers/filters is rather questionable -- since any of them can be disabled, checking the version is a very bad way to test for their presence) -- Anton Khirnov _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
