[mb-users] Automated tag update checking

[Kuno Woudt]

On Tue, Oct 16, 2007 at 04:25:55PM +0200, Age Bosma wrote:
> Mike Lerch wrote:
> >That got me thinking, you know what would be really useful?  Add a tag
> >(or maybe we already have it!) that records the vintage of the MB info
> >right in the file.  Then have Picard (or MusicIP or whatever) scan
> >through your collection to find files that the community has updated.
> >
> >Do other folks like that idea, or perhaps it already exists and I just
> >am missing it?
> >
> 
> I don't use Picard for tagging purposes but I believe that the 
> MusicBrainz track ID is stored in the tags if you use Picard to tag your 
> files. This means that the tracks you have are matched to the 
> corresponding track in MusicBrainz. If you lookup your tracks again you 
> will get an indication how much the tags match and you can choose to 
> re-tag your files again if changed have been made to that track in the 
> MB db.
> 
> So what you're looking for is already possible.

Picard is not really suitable for going through 'your collection'.  It
is useful on perhaps a dozen of albums, but if you throw your 100GB+
music collection at it it croaks.

What I would like to have, is a commandline tool [1] which I could run
automatically (once a month/week/day), to update the tags on those files
which have MBIDs.  The 'vintage of the MB info' as Mike calls it
wouldn't be useful currently, as there is no way to query MB on when
something related to an MBID has changed [2].  I think it would suffice
(for now) to look at the modification time of the files themselves, 
and do a query for each file which hasn't been changed in e.g. 1 month
(and compare the actual data to see if the file needs updating).

--Kuno.

[1] I have spent a little time on creating a commandline tool like
this, which obviously is not finished yet.  I would have preferred 
to use parts of picard for it, but it is not easy to seperate the 
parts that would be needed.  All parts of PicardQT seem very much 
bound to QT.  But it would be a shame to not be able to use features
like tagger scripts in such a commandline tool.

[2] correct me if i'm wrong.  It would be convenient if there was some
way to check how old data in MB is, preferably by just doing a HEAD
request to the webservice or something.  I believe this is also somehow
related to caching and allowing google to index MB, but again, I could 
be wrong.