#metabrainz

/

      • Freso
        And that release has that for the DJ mixer relationship.
      • 2016-03-16 07630, 2016

      • alastairp
      • 2016-03-16 07602, 2016

      • alastairp
        that has a <target-credit> (which, incidentally I don't think we support),
      • 2016-03-16 07616, 2016

      • alastairp
      • 2016-03-16 07659, 2016

      • Freso
      • 2016-03-16 07612, 2016

      • D4RK-PH0ENiX joined the channel
      • 2016-03-16 07643, 2016

      • D4RK-PH0ENiX has quit
      • 2016-03-16 07650, 2016

      • D4RK-PH0ENiX joined the channel
      • 2016-03-16 07652, 2016

      • alastairp
        sure, so that answers my initial question. but it seems that question is incorrect
      • 2016-03-16 07657, 2016

      • alastairp
        I don't want the value of that column
      • 2016-03-16 07627, 2016

      • alastairp
        ahhh. we're looking at relationships
      • 2016-03-16 07631, 2016

      • alastairp
        rather than attributes
      • 2016-03-16 07632, 2016

      • alastairp
        ah yes, what about work->composer->credited as ?
      • 2016-03-16 07624, 2016

      • alastairp
        ah no, that seems to be the same as the one you showed me
      • 2016-03-16 07618, 2016

      • alastairp
        ahhh, it seems it may only be for instrument relationships
      • 2016-03-16 07650, 2016

      • Freso
        Yay for consistency...
      • 2016-03-16 07618, 2016

      • ruaok
        woo, first pass off to chrisskye via Pipedrive.
      • 2016-03-16 07623, 2016

      • ruaok
        I like that system. :)
      • 2016-03-16 07635, 2016

      • alastairp
      • 2016-03-16 07635, 2016

      • ruaok
        we're all adult like now. :)
      • 2016-03-16 07646, 2016

      • alastairp
        so, for the name of the instrument in the credit
      • 2016-03-16 07649, 2016

      • alastairp
        not the name of the person
      • 2016-03-16 07601, 2016

      • alastairp
        sigh. I didn't want to have to dig into the database for this
      • 2016-03-16 07602, 2016

      • reosarevok
        ruaok: huh, we're using Estonian tech now? Nice :p
      • 2016-03-16 07633, 2016

      • ruaok
        I had no idea. Very good stuff.
      • 2016-03-16 07616, 2016

      • reosarevok
        Yup. Even have a dev office in town
      • 2016-03-16 07640, 2016

      • alastairp
        http://beta.musicbrainz.org/recording/f606f733-c1… piano [Yamaha and Steinway pianos]:
      • 2016-03-16 07644, 2016

      • alastairp
        that's what I was looking for
      • 2016-03-16 07655, 2016

      • alastairp
        Freso: thanks for pointing me to the other one, too
      • 2016-03-16 07637, 2016

      • chirlu
        ruaok: I remembered one thing that should be added to the schema change blog post.
      • 2016-03-16 07648, 2016

      • ruaok
        yeah?
      • 2016-03-16 07651, 2016

      • Freso
        alastairp: Np. :)
      • 2016-03-16 07657, 2016

      • chirlu
        We’ve decided to go from 10.04 LTS to 14.04 LTS as the minimum supported Ubuntu version.
      • 2016-03-16 07620, 2016

      • chirlu
        In particular relevant is the Perl version (5.20, I think, but will check).
      • 2016-03-16 07606, 2016

      • ruaok
        do you have a blog account? it would be very helpful if you could please make that edit.
      • 2016-03-16 07656, 2016

      • chirlu
        5.18, actually.
      • 2016-03-16 07600, 2016

      • chirlu
        No, I don’t.
      • 2016-03-16 07610, 2016

      • ruaok
        Would you like one? :)
      • 2016-03-16 07621, 2016

      • alastairp
        say no, say no!
      • 2016-03-16 07637, 2016

      • chirlu
        Well, I’ll probably need one one day …
      • 2016-03-16 07650, 2016

      • ruaok
        pm me your preferred email for that. :)
      • 2016-03-16 07617, 2016

      • ruaok
        invite on the way
      • 2016-03-16 07634, 2016

      • alastairp
        maybe someone can give me some ideas about this -
      • 2016-03-16 07640, 2016

      • alastairp
      • 2016-03-16 07650, 2016

      • alastairp
        to summarise, we have attributes which can be
      • 2016-03-16 07600, 2016

      • alastairp
        <attribute>founder</attribute>
      • 2016-03-16 07607, 2016

      • alastairp
        <attribute type="Key">G-flat major</attribute>
      • 2016-03-16 07614, 2016

      • alastairp
        <attribute value="BuxWV 1">number</attribute>
      • 2016-03-16 07635, 2016

      • alastairp
        <attribute credited-as="Yamaha and Steinway pianos">piano</attribute>
      • 2016-03-16 07613, 2016

      • alastairp
        the 2nd one can mean key=G-flat major, the 3rd number=BuxWV 1
      • 2016-03-16 07629, 2016

      • alastairp
        the 4th, perhaps piano (Yamaha and Steinway pianos), but not piano=Yamaha and Steinway pianos
      • 2016-03-16 07615, 2016

      • alastairp
        if you were to normalise this into a dictionary, with two keys, what would those keys be? (the first example would have just one of the keys)
      • 2016-03-16 07620, 2016

      • Freso
        I think both should be exposed so users can decide themselves whether they want to present as "piano", "Y&S pianos", or "piano (Y&S pianos)".
      • 2016-03-16 07629, 2016

      • Freso
        Or some entirely different way.
      • 2016-03-16 07641, 2016

      • alastairp
        yes, the plan is to expose it all
      • 2016-03-16 07658, 2016

      • alastairp
        the question is what should the keys of a dictionary be
      • 2016-03-16 07629, 2016

      • Freso
        instrument: piano, instrument-credit: Y&S pianos?
      • 2016-03-16 07646, 2016

      • alastairp
        this is an <attribute> element
      • 2016-03-16 07622, 2016

      • alastairp
        I don't want different keys depending on if it's a performance rel attribute, or a work series attribute, or a work attribute (if at all possible)
      • 2016-03-16 07619, 2016

      • Freso
        attribute[]['piano']['credited-as'] = u"Y&S pianos"?
      • 2016-03-16 07638, 2016

      • alastairp
        possibly
      • 2016-03-16 07647, 2016

      • alastairp
        I don't think piano should be a dict key though
      • 2016-03-16 07642, 2016

      • Freso
        attribute[] = (u'piano', credited-as: u'Y&S')?
      • 2016-03-16 07644, 2016

      • UmkaDK
        chirlu: Am I reading this right, as of next schema release minimum requirements will be updated to PostgreSQL 9.5 and perl 5.18?
      • 2016-03-16 07655, 2016

      • Freso
        UmkaDK: Yes.
      • 2016-03-16 07608, 2016

      • chirlu
        Yes.
      • 2016-03-16 07609, 2016

      • UmkaDK
        Awesome! :) Thanks Freso!!
      • 2016-03-16 07617, 2016

      • UmkaDK
        Thanks!
      • 2016-03-16 07618, 2016

      • alastairp
        Freso: what type of object is that?
      • 2016-03-16 07626, 2016

      • Freso
        alastairp: Pseudo-code object.
      • 2016-03-16 07651, 2016

      • ruaok
        chrisskye: I've added two users of our live data feed to pipedrive. I hope I didn't mess things up too badly for you.
      • 2016-03-16 07656, 2016

      • ruaok
        let me know if you need more info on these
      • 2016-03-16 07650, 2016

      • Freso
        ruaok: If you have the time, I think https://community.metabrainz.org/t/writer-publish… could benefit of a reply from you, as the founder of MB.
      • 2016-03-16 07617, 2016

      • ruaok
        on my list, but prod me later today, please
      • 2016-03-16 07622, 2016

      • Freso
        (Someone wants MB to store information for who to pay royalties to.)
      • 2016-03-16 07624, 2016

      • Freso
        Alright. :)
      • 2016-03-16 07606, 2016

      • ruaok
        yes, but that is generally a tough issue.
      • 2016-03-16 07623, 2016

      • ruaok
        because what if money flows into the wrong hands because of bad data in MB?
      • 2016-03-16 07640, 2016

      • ruaok
        what if people start wrangling the data in MB because it is used to determine payments?
      • 2016-03-16 07629, 2016

      • ruaok
        I'd love to have a more detailed conversation with this person, but it will take some of my time do that.
      • 2016-03-16 07645, 2016

      • Freso
        Sure. And I know it's not an easy thing, but I'd felt better you participating on that forum conversation, than me doing it, since pretty much all I know about it I've gleaned from writings (or conversations etc.) from you.
      • 2016-03-16 07616, 2016

      • ruaok
        understood. this is starting to get into the nuanced guts of the industry. :)
      • 2016-03-16 07631, 2016

      • jaron has quit
      • 2016-03-16 07633, 2016

      • chirlu
      • 2016-03-16 07600, 2016

      • Freso
      • 2016-03-16 07632, 2016

      • Jormangeud has quit
      • 2016-03-16 07648, 2016

      • Jormangeud joined the channel
      • 2016-03-16 07640, 2016

      • Jormangeud has quit
      • 2016-03-16 07656, 2016

      • Jormangeud joined the channel
      • 2016-03-16 07605, 2016

      • jaron joined the channel
      • 2016-03-16 07650, 2016

      • ruaok
        chirlu: looks ok. one bit of blogging etiquette I would like us to follow: If we amend a blog post (not to fix my usual crap typos), but make content changes, we should add a UPDATED: section at the bottom.
      • 2016-03-16 07611, 2016

      • ruaok
        UPDATED: Added a paragraph about supported Ubuntu versions.
      • 2016-03-16 07617, 2016

      • ruaok
        would be great if you could please add that.
      • 2016-03-16 07618, 2016

      • chirlu
        Well, I did? Not UPDATED, though.
      • 2016-03-16 07636, 2016

      • chirlu
        “(Edited 2016-03-16 at 13:55 UTC to add the upgraded Ubuntu requirement.)”
      • 2016-03-16 07655, 2016

      • chirlu
        Is it too inconspicuous? :-)
      • 2016-03-16 07630, 2016

      • ruaok
        ah, sorry, I didn't see that, the italics threw me off and I thought that was trailer guff already.
      • 2016-03-16 07637, 2016

      • ruaok
        that works. :)
      • 2016-03-16 07643, 2016

      • ruaok
        thanks!
      • 2016-03-16 07626, 2016

      • chirlu notices that the time is actually wrong …
      • 2016-03-16 07622, 2016

      • chirlu
        Oh, those time zones …
      • 2016-03-16 07623, 2016

      • alastairp
        If I'm making lists of tags of releases and release groups, does it make sense to combine them all together within a releasegroup?
      • 2016-03-16 07633, 2016

      • alastairp
        consider that the end goal is to only select tags that look like genres
      • 2016-03-16 07643, 2016

      • alastairp
        so things like 'reissue' won't be a problem
      • 2016-03-16 07615, 2016

      • Freso
        alastairp: hat would be the goal of combining them together?
      • 2016-03-16 07616, 2016

      • alastairp
        e.g. if an RG has 2 releases, and the rg and each release has distinct tags, merge them together
      • 2016-03-16 07623, 2016

      • alastairp
        to get more data
      • 2016-03-16 07645, 2016

      • chrisskye
        ruaok: thanks for the adds into the pipeline; works like a charm
      • 2016-03-16 07651, 2016

      • alastairp
        e.g., if we have a recording on release 'a' in AB, but no tags for it, but we have tags on release `b`, which is in the same RG
      • 2016-03-16 07603, 2016

      • alastairp
        should we consider the tags?
      • 2016-03-16 07606, 2016

      • ruaok
        chrisskye: great. I have a few more to punt over later.
      • 2016-03-16 07611, 2016

      • Freso
        Wouldn't that become less data? (Ie., you don't have data that the tag was used on two different releases, but only that it's included somewhere within an RG.)
      • 2016-03-16 07636, 2016

      • alastairp
        the data would potentially become less "accurate"
      • 2016-03-16 07640, 2016

      • alastairp
        if that's what you mean
      • 2016-03-16 07612, 2016

      • chrisskye
        ruaok: I changed all the headers...I think they make general sense for what we’re trying to keep track of, but suggestions welcome
      • 2016-03-16 07639, 2016

      • Freso
        Ah. I think it's fine to take RG-level tags and consider them for all R's, but I wouldn't take one R-level tag and consider it for all Rs ithin the RG.
      • 2016-03-16 07650, 2016

      • ruaok
        chrisskye: seemed ok to me.
      • 2016-03-16 07657, 2016

      • Freso
        But maybe that is still fine.
      • 2016-03-16 07615, 2016

      • Freso
        The edge cases where it would cause an issue are probably extremely far between.
      • 2016-03-16 07640, 2016

      • Freso
        I can't think of any (real or thought up) scenario where it wouldn't be applicable.
      • 2016-03-16 07647, 2016

      • Freso
        So go for it, probably.
      • 2016-03-16 07609, 2016

      • alastairp
        right, this is why I added the caveat that the tags we're going to be considering are only genre tags
      • 2016-03-16 07618, 2016

      • Freso
        Yeah, but even then.
      • 2016-03-16 07632, 2016

      • alastairp
        so, while I agree that there are edgecase tags that could be weird, I don't think genres are as bad
      • 2016-03-16 07636, 2016

      • Freso
        Maybe one release has a bonus CD with death metal remixes of an pop-reggae album.
      • 2016-03-16 07644, 2016

      • alastairp
        perhaps old cd1/cd2 style releases
      • 2016-03-16 07645, 2016

      • alastairp
        yeah, that
      • 2016-03-16 07620, 2016

      • ruaok
        you ever get that moment, where you're sitting at your desk, listening to music and then you think "I should put on some music?"?
      • 2016-03-16 07620, 2016

      • alastairp
        I'll see what the number say. my gut feel is that this is small enough to not be an issue
      • 2016-03-16 07626, 2016

      • Freso
        (If anyone knows of a specific case of that, I would love to give it a listen. Really.)
      • 2016-03-16 07629, 2016

      • Freso
        alastairp: Yeah.
      • 2016-03-16 07610, 2016

      • jaron has quit
      • 2016-03-16 07628, 2016

      • outsidecontext joined the channel
      • 2016-03-16 07647, 2016

      • outsidecontext has left the channel
      • 2016-03-16 07602, 2016

      • outsidecontext joined the channel
      • 2016-03-16 07655, 2016

      • justharshal has quit
      • 2016-03-16 07600, 2016

      • manu-chroma joined the channel
      • 2016-03-16 07641, 2016

      • zas
        expect mail issues, we are doing the switch to gmail, please report anything that may be related
      • 2016-03-16 07612, 2016

      • JesseW joined the channel
      • 2016-03-16 07650, 2016

      • Jormangeud has quit
      • 2016-03-16 07617, 2016

      • Jormangeud joined the channel
      • 2016-03-16 07658, 2016

      • CallerNo6
        hey GSoC mentor-peoples, you wanted all the ideas pages to be 2016-specific, correct?
      • 2016-03-16 07636, 2016

      • ruaok
        Freso: post a long response to tdruth.
      • 2016-03-16 07653, 2016

      • ruaok
        I find it grand that at least one entertainment lawyer is smart enough to use us
      • 2016-03-16 07646, 2016

      • CallerNo6
        At least, that's what I thought was agreed upon. But some or all of you have been creating non-year-specific pages, i.e. wiki/Summer_of_Code/<sub-project>
      • 2016-03-16 07652, 2016

      • Jormangeud has quit
      • 2016-03-16 07627, 2016

      • Jormangeud joined the channel
      • 2016-03-16 07635, 2016

      • diana_olhovyk has quit
      • 2016-03-16 07611, 2016

      • Gentlecat
        CallerNo6: we can add redirects, right?
      • 2016-03-16 07656, 2016

      • CallerNo6
        Gentlecat, sure. So e.g. gsoc/2016/<project> could redirect to the last relevant edit of gsoc/<project>