On Mon, 6 Aug 2018 11:44:16 +0200 Landry Breuil wrote: > On Mon, Aug 06, 2018 at 11:14:18AM +0200, Charlène wrote: > > Hi! > > > > Here is an update for audio/abcde, that requires the updated > > www/p5-WebService-MusicBrainz and the new > > audio/p5-MusicBrainz-DiscID ports i've submitted earlier. > > > > We are currently several versions behind and the changelog is too > > huge to detail here [1]. The biggest changes: > > > > - It now requires Perl to access the musicbrainz.org v2.0 API, that > > has become the default over CDDB (which is still available) to > > fetch tracks informations. > > - MP3 encoding now requires audio/py-eyed3 with the default > > configuration file. Our port only supports the default OGG Vorbis > > encoding, so it hasn't been added to the Makefile as a RUN_DEPENDS, > > but DESCR has been updated accordingly. > > > > I also added a patch because abcde wasn't able to get tracks offsets > > from cdparanoia on OpenBSD and submitted a bug report to upstream > > [2] > > > > As usual, any comment is welcome! > > This is nice but how much has this been tested ? There were attempts > to update it in the past months, and it was busted (cf > https://marc.info/?l=openbsd-ports&m=151722260128733&w=2, > https://marc.info/?t=152149700300007&r=1&w=2) - a bug reporte was > opened upstream at > https://abcde.einval.com/bugzilla/show_bug.cgi?id=66.
What i've successfully tested, under chroot build and runtime indeed, but also on my own system: - Using a 2.7.2 default ~/.abcde.conf with a 2.9.2 build causes no issues - MusicBrainz works, and CDDB still works if you instruct abcde to use it instead - Encoding: vorbis, speex, flac, mp3, aiff. Standalone and combined ( -o ogg,mp3...) - `abcde 1 3-5` rips tracks 1, 3, 4, 5 as expected. Due to the vast amount of options available, everything cannot be tested, but if somebody has a feature not included here they need to test with that port update, and has no time to build it and its new dependencies, feel free to send me a test case! > I dunno if that's the same issue, and if your patch from > https://abcde.einval.com/bugzilla/show_bug.cgi?id=86 fixes it, but it > would be nicer to clarify how much testing you've done :) > Oh, i should search the archive next time! The root issue is similar, but it has not the same consequences, on 2.9.1 and 2.9.2 it's a bash math evaluation failure instead. My awk proposal to upstream *may* be more portable because it's more basic, but as the wording in my upstream bug report implies, i would like to hear from them about it. Charlène. PS: This is how a successful run looks from here - https://octodon.social/@julianaito/100499923240970547
