On 2018/08/28 21:11, Brian Callahan wrote: > > > On 8/28/2018 6:01 PM, Charlene Wendling wrote: > > On Tue, 28 Aug 2018 16:18:12 -0400 > > Brian Callahan wrote: > > > >> On 08/08/18 07:51, Charlène wrote: > >>> On Wed, 8 Aug 2018 09:17:52 +0100 > >>> Stuart Henderson wrote: > >>> > >>>> On 2018/08/07 20:27, Charlène wrote: > >>>>> On Mon, 6 Aug 2018 10:52:23 +0200 > >>>>> Charlène wrote: > >>>>> > >>>>>> Hi, > >>>>>> > >>>>>> I'm currently working on an update for audio/abcde, that requires > >>>>>> this update and a new port (audio/p5-MusicBrainz-DiscID). > >>>>>> > >>>>>> Abcde wants WebService::MusicBrainz v1.0.4, the latest release as > >>>>>> i'm writing this. > >>>>>> > >>>>>> Many things changed, i'm quoting upstream: > >>>>>> > >>>>>> « Version 1.0 and future releases are not backward compatible > >>>>>> with pre-1.0 releases. This is a complete re-write using version > >>>>>> 2.0 of the MusicBrainz API and Mojolicious. » > >>>>>> > >>>>>> So i've checked with sqlports, no other ports than the future > >>>>>> updated audio/abcde depends on this port. > >>>>>> > >>>>>> I'm also joining a very basic script, if you want to test the > >>>>>> module. > >>>>>> > >>>>>> Comments and OK are welcome! > >>>>>> > >>>>>> Charlène. > >>>>> Hi, > >>>>> > >>>>> I'm proposing myself as MAINTAINER for this port, so i'm joining a > >>>>> new diff. > >>>>> > >>>>> Charlène. > >>>> +LIB_DEPENDS= www/p5-Mojo>=7.13 > >>>> > >>>> LIB_DEPENDS is the wrong thing for Perl modules, they are > >>>> RUN_DEPENDS. > >>>> > >>>> LIB_DEPENDS without an associated WANTLIB is always wrong. > >>>> > >>>> > >>> Hi Stuart, > >>> > >>> Thanks for reviewing, i'm taking note of it and changed accordingly, > >>> also removed a spacing inconsistency on MAINTAINER line. > >>> > >>> I've retested those changes, it builds and runs fine on my amd64 > >>> machine. > >>> > >>> Charlène. > >> Some of the tests fail here though I'm not sure how much that matters. > >> Stuart, you can go commit the update if you want. ok for me. > >> ~Brian > >> > > Hi, > > > > I looked at the issue, Area.t fails at line 45 because the API returns > > something different than expected, even by the doc [1]. ext:score is not > > there, it's ns2:score now, see the live result here [2]. > > > > The final target and only consumer that would be an upgraded > > audio/abcde doesn't use area queries. If needed i'm joining an alternate > > version of the port with the patched test (i'll need it for upstream > > anyway): > > > > ===> Regression tests for p5-WebService-MusicBrainz-1.0.4 > > t/Area.t ....... ok > > t/Artist.t ..... ok > > t/Label.t ...... ok > > t/Recording.t .. ok > > t/Release.t .... ok > > All tests successful. > > Files=5, Tests=76, 38 wallclock secs ( 0.05 usr 0.08 sys + 1.99 cusr > > 0.54 csys = 2.66 CPU) Result: PASS > > > > Charlène. > > > > > > [1] > > https://musicbrainz.org/doc/Development/XML_Web_Service/Version_2/Search > > [2] http://musicbrainz.org/ws/2/area/?query=%22California%22 > > This is fine with me too. > > ~Brian >
OK.
