Re: [mb-style] RFC-327: Featured Artists
Am 12.07.2011, 16:45 Uhr, schrieb Andii Hughes gnu_and...@member.fsf.org: This deals with the legacy data issue (ws/1 still provides data in pre-NGS format) by retaining the existing guideline for track listings, while making use of artist credits at recording level. When ws/1 is finally removed, we can allow track listings to be entered as on the release, as some have suggested. Comments? there is a track and a recording lvl because some people crave normalisation others want as on cover. so thats how it should be, imho, and for compatibility reasons without artist credits on track lvl. i do think a switch to link the featured artist(s) to the title or the primary artist would be great. also the expamples are confusing to me. i'd just give examples on tracklvl and then on recording lvl. it was not clear to me that the indented entrys are for recording lvl, i related them to the Track-Level -- lorenz pressler PGP 0x92E9551A ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-331: Add CD Baby Relationship Type
I'm against any new relationship type that move away MusicBrainz from a neutral position regarding music sellers. A project claiming to be the open music encyclopedia can't be a link farm for music shops. MusicBrainz goal is not to be the universal way to find and buy your music. If this relationship type is accepted, then there's no reason to not add other relationship types for other music shops (and there are hundreds ones). And we will just end up with ton of spam links that will clutter the database and web site. When creating a new relationship type we should always ask us: What value does this bring to MusicBrainz? Here I don't see any. - Aurélien On 07/17/2011 11:58 PM, Johannes Weißl wrote: Hello, another proposal related to RFC-329: add a CD Baby Relationship Type. Here a short explanation why I think this type is useful and necessary: - CD Baby has artist pages now, to which we can't link very well (only via the generic can be purchased for download/mail-order at (which one should be used? Both?) - We currently have a supported has cover at CD Baby AR. This however doesn't express the fact that the music can bought there as well. As a consequence, I've seen releases that have three types of links (cover at, purched for download and purchased for mail-order) all to the same URL, which is quite silly, but a logical consequence of the missing AR. As I imagine it, the new AR would obsolete the other three ARs for CD Baby. This doesn't need to be done automatically, but of course can. So if a release gets linked via this AR to a CD Baby release page, the cover would get fetched automatically (like it is the case for Amazon.com). The proposed mappings are: MB artist- CD Baby artist page MB release- CD Baby release page Previous discussion: http://lists.musicbrainz.org/pipermail/musicbrainz-style/2011-July/012503.html Wiki page: http://wiki.musicbrainz.org/Proposal:CD_Baby_Relationship_Type Expiration date: Sun, 24 Jul 2011 22:00 UTC Johannes ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-331: Add CD Baby Relationship Type
On 07/18/2011 08:44 AM, Nicolás Tamargo de Eguren wrote: On Mon, Jul 18, 2011 at 9:35 AM, Aurélien Minoa.m...@free.fr wrote: I'm against any new relationship type that move away MusicBrainz from a neutral position regarding music sellers. A project claiming to be the open music encyclopedia can't be a link farm for music shops. MusicBrainz goal is not to be the universal way to find and buy your music. If this relationship type is accepted, then there's no reason to not add other relationship types for other music shops (and there are hundreds ones). Well, CD Baby is an affiliate that allows cover usage. I don't think a CD Baby relationship is worse than a has cover at CD Baby + a can be purchased for mail order at [CDBaby URL], as it turns two relationships into one. Half of the spam, I'd say! Why do you need the can be purchased for mail order at [CDBaby URL] relationship? The has cover at is enough. Furthermore you're forgetting that this proposal is more than just a release-url relationship. What the point of linking to an artist page on a shop? Affiliates - as you call them - are just a way to get cover art. 1. They're not bringing much money to MetaBrainz anyway 2. And once Rob's project of cover art archive will get out, we won't really need such partnership anymore - Aurélien ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-331: Add CD Baby Relationship Type
On Mon, Jul 18, 2011 at 9:54 AM, Aurélien Mino a.m...@free.fr wrote: On 07/18/2011 08:44 AM, Nicolás Tamargo de Eguren wrote: On Mon, Jul 18, 2011 at 9:35 AM, Aurélien Minoa.m...@free.fr wrote: I'm against any new relationship type that move away MusicBrainz from a neutral position regarding music sellers. A project claiming to be the open music encyclopedia can't be a link farm for music shops. MusicBrainz goal is not to be the universal way to find and buy your music. If this relationship type is accepted, then there's no reason to not add other relationship types for other music shops (and there are hundreds ones). Well, CD Baby is an affiliate that allows cover usage. I don't think a CD Baby relationship is worse than a has cover at CD Baby + a can be purchased for mail order at [CDBaby URL], as it turns two relationships into one. Half of the spam, I'd say! Why do you need the can be purchased for mail order at [CDBaby URL] relationship? The has cover at is enough. I don't need any of them, personally. I don't buy CDs anyway. I've seen people adding both though, because they want to show the album is for sale there, not that it has the cover there. Furthermore you're forgetting that this proposal is more than just a release-url relationship. What the point of linking to an artist page on a shop? As I said, I won't be using it, so I don't care. But if even the fairly restrictive and harsh What not to link to gave as only reason not to link to CD Baby artist pages that it is hard to find a permanent link for them instead of just they're not useful for MB, I gather some people do find them useful. Affiliates - as you call them - are just a way to get cover art. 1. They're not bringing much money to MetaBrainz anyway 2. And once Rob's project of cover art archive will get out, we won't really need such partnership anymore Even though I am fairly sure we'll be getting the archive at some point, I'd rather work with what we have now, and not with what we will have at an indefinite future point (if I'm not mistaken, not only coding for it hasn't started, but we don't even have the full approval of the MB Foundation yet). - Aurélien ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style -- Nicolás Tamargo de Eguren ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-331: Add CD Baby Relationship Type
Hello Aurélien, On Mon, Jul 18, 2011 at 08:54:20AM +0200, Aurélien Mino wrote: Why do you need the can be purchased for mail order at [CDBaby URL] relationship? The has cover at is enough. Well, but can be purchased links are not wrong, even if has cover at exists, aren't they? So people add them, even when the has cover at link exists. I've cleanup up a few releases like that, but I can understand users that are not satisfied with only has cover at. This new type would solve this issue! Furthermore you're forgetting that this proposal is more than just a release-url relationship. What the point of linking to an artist page on a shop? Since CD Baby is mostly used by independent artists, you can see it as the artists official online shop page (not just a random shop that sells music). Just like these artists have a myspace/facebook page to advertise, they have a CD Baby page to sell their music. The bonding between CD Baby and the artists is much stronger than for example Amazon.com. Many users (and sometimes even artists!) think that CD Baby is the label for a release. Johannes ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] CSG: Classical superworks and performance of links
2011/7/18, Nicolás Tamargo de Eguren reosare...@gmail.com: I am looking at http://musicbrainz.org/work/51ef99ac-488d-4c35-bef3-fc25bf73ca39 (Divertissement, Op. 38 by Fernando Sor). We have a recording of the full work in one go, and a recording of each of its two movements. Should the full recording be linked to both movements, or to the full work as it is now? To the full Work IMO. I understand linking to both movements would mean the same, but linking to each movement would seem unnecessarily complicated to me, and I don't see any benefit in doing so. Imagine a full opera in one take... -- Frederic Da Vitoria (davitof) Membre de l'April - « promouvoir et défendre le logiciel libre » - http://www.april.org ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-331: Add CD Baby Relationship Type
Aurélien Mino wrote: Why do you need the can be purchased for mail order at [CDBaby URL] relationship? The has cover at is enough. It depends what your aim is. If you want to say that you can purchase the release for both mail order and download, and there's also cover art, then it's logical to want to add multiple relationships. If you just want to add a link to CD Baby then it's logical to add only one and not care about the exact relationship type. IMO the cover art relationship was/is the wrong relationship to use anyway. It's not a link to an image. What the point of linking to an artist page on a shop? Why were download/mail-order relationships ever added at artist level if we don't want to allow them there? Nikki ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On 18 July 2011 07:09, lorenz pressler l...@gmx.at wrote: Am 12.07.2011, 16:45 Uhr, schrieb Andii Hughes gnu_and...@member.fsf.org: This deals with the legacy data issue (ws/1 still provides data in pre-NGS format) by retaining the existing guideline for track listings, while making use of artist credits at recording level. When ws/1 is finally removed, we can allow track listings to be entered as on the release, as some have suggested. Comments? there is a track and a recording lvl because some people crave normalisation others want as on cover. so thats how it should be, imho, and for compatibility reasons without artist credits on track lvl. How do you mean? Why would you disallow artist credits on track level? i do think a switch to link the featured artist(s) to the title or the primary artist would be great. They already are part of the title. The proposal is to allow them to be moved to the artist for recordings. also the expamples are confusing to me. i'd just give examples on tracklvl and then on recording lvl. it was not clear to me that the indented entrys are for recording lvl, i related them to the Track-Level That's exactly how it is now. The first line is the track level, then the recording level is the indented list. I can't see a much better way of presenting it. -- lorenz pressler PGP 0x92E9551A ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style -- Andii :-) ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
There seem to be several unresolved questions here we should try to address separately: 1. How does as on cover (on track/release level) *ideally* translate into NGS structures? For an album by artist Foo with a track Quux featuring artist Bar: A) [track] Quux (feat. Bar) [Artist Credit] Foo (as we had pre-NGS) or B) [track] Quux [Artist Credit] Foo feat. Bar (using the NGS multiple artist feature, with the link phrase exactly as on cover) ? 2. How should the featured artists *ideally* be treated on the (normalized) recording level? Assuming there is consensus that feat.-information should not be part of the title field at recording level: A) [Artist Credit] Foo Bar (as by default) B) [Artist Credit] Foo feat. Bar (assuming feat. has a specific different meaning, not to be normalized to ) 3. Do we want/need a delay on application of (part of) these *ideal* rules, once agreed upon, because of the ws/1 issue? 4. Do any or all of the above rules apply only to the exact wording feat. (or featuring, features) or would it also affect (how?) other link terms (in other languages) like with, en duo avec, con The guideline should IMO reflect the ideal regulations, with a warning box for not yet to be used features because of ws/1 compatibility issues, which could be removed (without further debate) when it's no longer used. Chris/chabreyflint ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On 18 July 2011 10:19, SwissChris swissch...@gmail.com wrote: There seem to be several unresolved questions here we should try to address separately: 1. How does as on cover (on track/release level) ideally translate into NGS structures? For an album by artist Foo with a track Quux featuring artist Bar: A) [track] Quux (feat. Bar) [Artist Credit] Foo (as we had pre-NGS) or B) [track] Quux [Artist Credit] Foo feat. Bar (using the NGS multiple artist feature, with the link phrase exactly as on cover) ? Either or neither may be as on the cover, because we don't know what's on the cover. 2. How should the featured artists ideally be treated on the (normalized) recording level? Assuming there is consensus that feat.-information should not be part of the title field at recording level: A) [Artist Credit] Foo Bar (as by default) B) [Artist Credit] Foo feat. Bar (assuming feat. has a specific different meaning, not to be normalized to ) The second is correct. The first is a collaboration, not a featured artist. 3. Do we want/need a delay on application of (part of) these ideal rules, once agreed upon, because of the ws/1 issue? It seems I'm alone in wanting track list changes delayed, even though there is still very little NGS support. 4. Do any or all of the above rules apply only to the exact wording feat. (or featuring, features) or would it also affect (how?) other link terms (in other languages) like with, en duo avec, con We should normalise to feat. where appropriate at recording level. The guideline should IMO reflect the ideal regulations, with a warning box for not yet to be used features because of ws/1 compatibility issues, which could be removed (without further debate) when it's no longer used. That sounds sensible, but I'm not sure if it wouldn't be too confusing. I also have no idea what's wanted as track level; just 'follow the cover'? Chris/chabreyflint ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style -- Andii :-) ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-330: Add Allmusic Relationship Type
I don't really see the point in shortening the URLs to just p4147. We don't do it for most sites and I'd rather see the full URL on the relationships tab and Allmusic in the sidebar. I don't think the first guideline is necessary either, we can just add the URL formats to the list on http://wiki.musicbrainz.org/Style/Relationships/URLs#Standardised_URLs Adding some Javascript to clean them up would be trivial (and any that get through anyway will get cleaned up eventually by someone like me). One other thing, www.allmusic.com or just allmusic.com? Both work without redirecting... Nikki Johannes Weißl wrote: Hi there! I propose to add an Allmusic Relationship Type. Since the general opinion seems to be that the reasons not to link to allmusic.com are not valid anymore (see previous discussion), we need a proper relationship to actually be able to link to it. The proposed mappings are: MB artist - AMG artist page (artist/p...) MB release group - AMG album page (main entry) (album/r...) MB work - AMG song/work page (work/c... and song/t...) MB recording - AMG performance (performance/f...) Previous discussion: http://lists.musicbrainz.org/pipermail/musicbrainz-style/2011-July/012503.html Wiki page: http://wiki.musicbrainz.org/Proposal:Allmusic_Relationship_Type Expiration date: Sun, 24 Jul 2011 20:00 UTC Johannes ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] CSG: Classical superworks and performance of links
On Mon, 18 Jul 2011 10:15:03 +0200, Frederic Da Vitoria davito...@gmail.com wrote: 2011/7/18, Nicolás Tamargo de Eguren reosare...@gmail.com: I am looking at http://musicbrainz.org/work/51ef99ac-488d-4c35-bef3-fc25bf73ca39 (Divertissement, Op. 38 by Fernando Sor). We have a recording of the full work in one go, and a recording of each of its two movements. Should the full recording be linked to both movements, or to the full work as it is now? To the full Work IMO. I understand linking to both movements would mean the same, but linking to each movement would seem unnecessarily complicated to me, and I don't see any benefit in doing so. Imagine a full opera in one take... In this case I think it doesn't matter. Otherwise I agree with Calvin ( http://lists.musicbrainz.org/pipermail/musicbrainz-style/2011-June/012280.html ) that the most accurate way is to link to all performed movements. A try to interpret the results of that discussion: Recording linked to the relevant works (parts): accurate Recording linked to the super-work: fuzzy - a performance of the whole super-work, but unsure about exactly which parts are performed Recording linked to the super-work with partial performance AR: not identified - a performance of one or more parts (or part of a subpart) /symphonick ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On Mon, Jul 18, 2011 at 11:26 AM, Andii Hughes gnu_and...@member.fsf.orgwrote: On 18 July 2011 10:19, SwissChris swissch...@gmail.com wrote: There seem to be several unresolved questions here we should try to address separately: 1. How does as on cover (on track/release level) ideally translate into NGS structures? For an album by artist Foo with a track Quux featuring artist Bar: A) [track] Quux (feat. Bar) [Artist Credit] Foo (as we had pre-NGS) or B) [track] Quux [Artist Credit] Foo feat. Bar (using the NGS multiple artist feature, with the link phrase exactly as on cover) ? Either or neither may be as on the cover, because we don't know what's on the cover. 2. How should the featured artists ideally be treated on the (normalized) recording level? Assuming there is consensus that feat.-information should not be part of the title field at recording level: A) [Artist Credit] Foo Bar (as by default) B) [Artist Credit] Foo feat. Bar (assuming feat. has a specific different meaning, not to be normalized to ) The second is correct. The first is a collaboration, not a featured artist. 3. Do we want/need a delay on application of (part of) these ideal rules, once agreed upon, because of the ws/1 issue? It seems I'm alone in wanting track list changes delayed, even though there is still very little NGS support. 4. Do any or all of the above rules apply only to the exact wording feat. (or featuring, features) or would it also affect (how?) other link terms (in other languages) like with, en duo avec, con We should normalise to feat. where appropriate at recording level. The guideline should IMO reflect the ideal regulations, with a warning box for not yet to be used features because of ws/1 compatibility issues, which could be removed (without further debate) when it's no longer used. That sounds sensible, but I'm not sure if it wouldn't be too confusing. I also have no idea what's wanted as track level; just 'follow the cover'? Well, this is exactly my first point: If the cover spells a track as Quux (feat. Bar) should follow the cover be: A) [track] Quux (feat. Bar) [Artist Credit] Foo (as we had pre-NGS) or B) [track] Quux [Artist Credit] Foo feat. Bar (using the NGS multiple artist feature, with the link phrase exactly as on cover) ? Chris/chabreyflint ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style -- Andii :-) ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On 18 July 2011 10:33, SwissChris swissch...@gmail.com wrote: On Mon, Jul 18, 2011 at 11:26 AM, Andii Hughes gnu_and...@member.fsf.org wrote: On 18 July 2011 10:19, SwissChris swissch...@gmail.com wrote: There seem to be several unresolved questions here we should try to address separately: 1. How does as on cover (on track/release level) ideally translate into NGS structures? For an album by artist Foo with a track Quux featuring artist Bar: A) [track] Quux (feat. Bar) [Artist Credit] Foo (as we had pre-NGS) or B) [track] Quux [Artist Credit] Foo feat. Bar (using the NGS multiple artist feature, with the link phrase exactly as on cover) ? Either or neither may be as on the cover, because we don't know what's on the cover. 2. How should the featured artists ideally be treated on the (normalized) recording level? Assuming there is consensus that feat.-information should not be part of the title field at recording level: A) [Artist Credit] Foo Bar (as by default) B) [Artist Credit] Foo feat. Bar (assuming feat. has a specific different meaning, not to be normalized to ) The second is correct. The first is a collaboration, not a featured artist. 3. Do we want/need a delay on application of (part of) these ideal rules, once agreed upon, because of the ws/1 issue? It seems I'm alone in wanting track list changes delayed, even though there is still very little NGS support. 4. Do any or all of the above rules apply only to the exact wording feat. (or featuring, features) or would it also affect (how?) other link terms (in other languages) like with, en duo avec, con We should normalise to feat. where appropriate at recording level. The guideline should IMO reflect the ideal regulations, with a warning box for not yet to be used features because of ws/1 compatibility issues, which could be removed (without further debate) when it's no longer used. That sounds sensible, but I'm not sure if it wouldn't be too confusing. I also have no idea what's wanted as track level; just 'follow the cover'? Well, this is exactly my first point: If the cover spells a track as Quux (feat. Bar) should follow the cover be: A) [track] Quux (feat. Bar) [Artist Credit] Foo (as we had pre-NGS) or B) [track] Quux [Artist Credit] Foo feat. Bar (using the NGS multiple artist feature, with the link phrase exactly as on cover) ? As I said above, it depends what the cover says! It may say a, it may say b or it may say something else, C. Chris/chabreyflint ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style -- Andii :-) ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style -- Andii :-) ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] CSG: Classical superworks and performance of links
2011/7/18, symphonick symphon...@gmail.com: Isn't this what the partial attribute is for? Or do you mean the editor doesn't even know if the performance is partial or not? But then, when ARing to a movement, how are we sure it is a full performance of the movement? -- Frederic Da Vitoria (davitof) Membre de l'April - « promouvoir et défendre le logiciel libre » - http://www.april.org ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] CSG: Classical superworks and performance of links
On Mon, 18 Jul 2011 11:43:33 +0200, Frederic Da Vitoria davito...@gmail.com wrote: 2011/7/18, symphonick symphon...@gmail.com: Recording linked to the super-work: fuzzy - a performance of the whole super-work, but unsure about exactly which parts are performed Isn't this what the partial attribute is for? Or do you mean the editor doesn't even know if the performance is partial or not? But then, when ARing to a movement, how are we sure it is a full performance of the movement? Note that it's recording - super-work, not recording-movement. I meant that it's presented as a full performance of a super-work, say Händel's Messiah, but we don't know exactly which parts; different versions of arias, maybe some movements left out, different arrangements and so on. /symphonick ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
Hello, On 18/07/11 12:07, SwissChris wrote: When cover says Track (feat. Artist) should this (follow the cover) translate into Track (feat. artist) (on track field) [by] Main Artist (on AC field) or (as I think is the only reasonable way) as Track (on track field) [by] Main Artist feat. featured artist (on AC field). I don't think we're coming to anything productive without answering this question first! Here are two examples (warning, large images :). Track 14 on Mysteryland 2010 is listed on the cover as: track title: Born Again artist credits: RICKY L FEAT. M:CK http://www.frob.nl/mb/2011/mysteryland.black.2010.jpg Track 1 on BoA is listed on the cover as: track title: I Did It For Love (featuring Sean Garret) artist credits: BoA http://www.frob.nl/mb/2011/smusa01.jpg I would prefer that we do not move Sean Garret in the second example to the artist credits of a track --- however, I wouldn't veto such a proposal as I can understand that many people would think it more important to properly link an artist to an artist entry in our database than it is to precisely follow the order of things as they are on the cover. -- kuno / warp. ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
Hello, On 18/07/11 12:14, Andii Hughes wrote: That's the whole point of starting with a (mostly) blank slate for parts which are largely new for NGS. To not impose new guidelines without going through the usual process. And something has to hold while new guidelines are being prepared. To say these interim guidelines are anything other than the previously approved ones is imposing new ones. If you think so, then so be it, we've imposed new guidelines. We've worked on the NGS guidelines for a year before the NGS release. During that year we've had many discussions on how to deal with these new entities. At various points throughout the process we've asked for feedback and help, through this mailinglist and on irc. If you're unhappy with the approach we've chosen, it's a bit late too complain about it now, there was more than enough time to give us that feedback before the NGS release. -- kuno / warp. ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] CSG: Classical superworks and performance of links
2011/7/18, symphonick symphon...@gmail.com: On Mon, 18 Jul 2011 11:43:33 +0200, Frederic Da Vitoria davito...@gmail.com wrote: 2011/7/18, symphonick symphon...@gmail.com: Recording linked to the super-work: fuzzy - a performance of the whole super-work, but unsure about exactly which parts are performed Isn't this what the partial attribute is for? Or do you mean the editor doesn't even know if the performance is partial or not? But then, when ARing to a movement, how are we sure it is a full performance of the movement? Note that it's recording - super-work, not recording-movement. I meant that it's presented as a full performance of a super-work, say Händel's Messiah, but we don't know exactly which parts; different versions of arias, maybe some movements left out, different arrangements and so on. Yes, but do we know it that often for movement tracks? Some performers remove some parts of works (of movements) like repeats, and most listeners don't notice it. Many classical releases don't bother to mention all the tiny details (I never saw written that a performer had omitted a repeat) so that you can only know for sure if you have the score(s) (or a good memory). I agree we probably miss something here, between the full, the unknown and the partial performances, but I feel we are considering as full many performances which, while not really partial (90% of the notes are there) are not quite full either. So either we state that the reverse of partial is actually almost full or apparently full, or we should add a full attribute to performance ARs and only users who have the scores (or an excellent musical memory) should check this attribute. I believe the first solution is much more reasonable :-) -- Frederic Da Vitoria (davitof) Membre de l'April - « promouvoir et défendre le logiciel libre » - http://www.april.org ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
2011/7/18, Kuno Woudt k...@frob.nl: Hello, On 18/07/11 12:14, Andii Hughes wrote: That's the whole point of starting with a (mostly) blank slate for parts which are largely new for NGS. To not impose new guidelines without going through the usual process. And something has to hold while new guidelines are being prepared. To say these interim guidelines are anything other than the previously approved ones is imposing new ones. If you think so, then so be it, we've imposed new guidelines. We've worked on the NGS guidelines for a year before the NGS release. During that year we've had many discussions on how to deal with these new entities. At various points throughout the process we've asked for feedback and help, through this mailinglist and on irc. If you're unhappy with the approach we've chosen, it's a bit late too complain about it now, there was more than enough time to give us that feedback before the NGS release. Like Andii, I would NOT use un-normalized data. I'd even be more extreme, I'd entirely remove things like feat which don't make much sense to me. From my point of view, an artist either performed or he did not, period. BUT - one frequent issue with edits before NGS was fights beween pro-normalization and pro-as-printed - one of the proposed benefits of NGS was allowing to enter data as printed as well as normalized. So that now I really don't know how I'd explain to a pro-as-printed: ok, we sold you NGS as solving all the issues, now NGS is here, it is much more clumsy to use than pre-NGS, there are almost twice as much fields, but the fields where you can enter data as printed are.. well, nowhere. I agree that entering data in 2 conflicting ways is going to be complicated. But if we don't, then I really fail to see what track titles and release titles are for. Let's do with recording titles and release group titles only, and we'll be perfectly normalized. I am quite serious here, let's remove track and release titles entirely, I won't complain, recording titles and release group titles are probably all I will ever use and data input will obviously be much simpler. And don't try to sell me a 2-level normalization system. I already had difficulties to memorize a few of the mysterious rules of the pre-NGS one-level system, I'd never manage to learn a 2-level system. -- Frederic Da Vitoria (davitof) Membre de l'April - « promouvoir et défendre le logiciel libre » - http://www.april.org ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] CSG: Classical superworks and performance of links
On Mon, 18 Jul 2011 13:29:56 +0200, Frederic Da Vitoria davito...@gmail.com wrote: 2011/7/18, symphonick symphon...@gmail.com: Recording linked to the super-work: fuzzy - a performance of the whole super-work, but unsure about exactly which parts are performed I agree we probably miss something here, between the full, the unknown and the partial performances I don't know if we agree or disagree? Let me try again: 1. A performance AR between a recording the appropriate movements is the most accurate we can do. Symphony No. 40 in G minor is a performance of Sinfonie in g K. 550: I. Molto Allegro (2. Fassung) Symphony No. 40 in G minor is a performance of Sinfonie in g K. 550: II. Andante (2. Fassung) 2. The accuracy of a performance AR between a recording a super-work is depending on the parts linked to the super-work. Symphony No. 40 in G minor is a performance of Sinfonie in g K. 550 Which version? (Works structure isn't clearly defined yet, this super-work could contain 2 superworks or 8 movements. But the issue regarding accuracy remains.) /symphonick ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On 18 July 2011 12:06, Kuno Woudt k...@frob.nl wrote: Hello, On 18/07/11 12:14, Andii Hughes wrote: That's the whole point of starting with a (mostly) blank slate for parts which are largely new for NGS. To not impose new guidelines without going through the usual process. And something has to hold while new guidelines are being prepared. To say these interim guidelines are anything other than the previously approved ones is imposing new ones. If you think so, then so be it, we've imposed new guidelines. We've worked on the NGS guidelines for a year before the NGS release. During that year we've had many discussions on how to deal with these new entities. At various points throughout the process we've asked for feedback and help, through this mailinglist and on irc. If you're unhappy with the approach we've chosen, it's a bit late too complain about it now, there was more than enough time to give us that feedback before the NGS release. Right, you've lost me now. There was a discussion of featured artists (which I took part in) before NGS was released. It never went to a vote: http://lists.musicbrainz.org/pipermail/musicbrainz-style/2011-April/011250.html which I presume means that the existing status quo is retained. The only guideline available is: http://wiki.musicbrainz.org/Style/Recording_and_release_group_titles/Featured_artists and now you're trying to imply that this somehow doesn't apply in some cases. Can you please point me to these other guidelines on featured artists if the linked one does not apply for track listings? Certainly no-one seems to be aware of them at present, which is why there is all this confusion. -- kuno / warp. ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style -- Andii :-) ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
There were two contrary statements about ws/1: Andii Hughes wrote: So again: Is there any chance to fix it? E.g. by letting ws/1 automatically merge all credited artists into one collaboration artist. Or (probably better) by using only the first credited artist as the track artist and automatically appending all others to the track title using the join phrases. See my earlier e-mail about ws/1. I would support the server being altered to automatically move feat. credits to the title and not have it in the guideline, but this isn't what it does at present. Calvin Walton-2 wrote: On Fri, 2011-07-15 at 17:16 -0700, Paul C. Bryan wrote: 3. Ideally, /ws/1 should handle the feat. link phrase by expressing the AC as ETI of the tracks it returns. Unfortunately, this is impossible to do in any sort of general way, mostly because the artist credits are used for more than just featured artists. Even if you limit it to just 'feat.' link phrases, there are several artists in the database that it would miss, and several collaborations or special artist credits that it would rewrite inappropriately. The current ws/1 method of simply making up a fake artist with a name that contains the entire artist credit is really the best we can do :/ Is it possible and does it make sense to change ws/1 in the described way or not? Wouldn't this always produce the same results for featured artists as it did in pre-NGS times? Even if it is extended to other link phrases like 'with', 'avec', 'con'? (But leaving out '', 'and', '+' and the like and possibly also 'vs.' for which all credited artists should probably remain in the artist field.) And BTW, what does ws/1 do about correctly entered collaborations like e.g. http://musicbrainz.org/release/58bad121-bfab-4dda-89f8-4b1bc092de44 ? I suppose these are just as ws/1 incompatible as featured artists but ws/1 has to live with them just as it has to with multi disc releases. -- View this message in context: http://musicbrainz.1054305.n4.nabble.com/RFC-327-Featured-Artists-tp3662499p3675138.html Sent from the Musicbrainz - Style mailing list archive at Nabble.com. ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
Hello, On 18/07/11 14:17, Andii Hughes wrote: Can you please point me to these other guidelines on featured artists if the linked one does not apply for track listings? Certainly no-one seems to be aware of them at present, which is why there is all this confusion. http://wiki.musicbrainz.org/Style/Track_and_release_titles ^ that is the entire guideline which applies to track and release titles. There is no Featured artists guideline for track and release titles. -- kuno / warp. ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On 18 July 2011 13:32, Kuno Woudt k...@frob.nl wrote: Hello, On 18/07/11 14:17, Andii Hughes wrote: Can you please point me to these other guidelines on featured artists if the linked one does not apply for track listings? Certainly no-one seems to be aware of them at present, which is why there is all this confusion. http://wiki.musicbrainz.org/Style/Track_and_release_titles ^ that is the entire guideline which applies to track and release titles. There is no Featured artists guideline for track and release titles. That's just the old release title guideline. What happens with the artist credit? -- kuno / warp. ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style -- Andii :-) ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
swisschris wrote: There seem to be several unresolved questions here we should try to address separately: 1. How does as on cover (on track/release level) *ideally* translate into NGS structures? For an album by artist Foo with a track Quux featuring artist Bar: A) [track] Quux (feat. Bar) [Artist Credit] Foo (as we had pre-NGS) or B) [track] Quux [Artist Credit] Foo feat. Bar (using the NGS multiple artist feature, with the link phrase exactly as on cover) ? B. 'As printed' does IMO not mean that additional artist info has to be part of the track title if the cover happens to have them next to each other. Otherwise we'd end up with A for all 'feat.' tracks on single artist releases and with B for almost all occurrences of the same track on VA compilations. 2. How should the featured artists *ideally* be treated on the (normalized) recording level? Assuming there is consensus that feat.-information should not be part of the title field at recording level: A) [Artist Credit] Foo Bar (as by default) B) [Artist Credit] Foo feat. Bar (assuming feat. has a specific different meaning, not to be normalized to ) Usually, '' and 'feat.' have different meanings and shouldn't be merged into one normalized phrase. But we could allow only a couple of standardized join phrases and make them selectable from a dropdown. E.g.: '' as a normalized expression for '', 'and', '+', ... 'with' as a normalized expression for 'with', 'avec', 'con', 'mit', ... 'duet with' as a normalized expression for 'duet with', 'Duett mit', 'en duo avec' 'vs.' ... 3. Do we want/need a delay on application of (part of) these *ideal* rules, once agreed upon, because of the ws/1 issue? Not if there is a reasonable workaround for the current ws/1 issue. But it is currently not clear if featured artist credits can automatically be moved from the artist field into ETI. 4. Do any or all of the above rules apply only to the exact wording feat. (or featuring, features) or would it also affect (how?) other link terms (in other languages) like with, en duo avec, con This guideline should ideally cover all cases where multiple artists are credited for one track or release. -- View this message in context: http://musicbrainz.1054305.n4.nabble.com/RFC-327-Featured-Artists-tp3662499p3675218.html Sent from the Musicbrainz - Style mailing list archive at Nabble.com. ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] CSG: Classical superworks and performance of links
2011/7/18, symphonick symphon...@gmail.com: On Mon, 18 Jul 2011 13:29:56 +0200, Frederic Da Vitoria davito...@gmail.com wrote: 2011/7/18, symphonick symphon...@gmail.com: Recording linked to the super-work: fuzzy - a performance of the whole super-work, but unsure about exactly which parts are performed I agree we probably miss something here, between the full, the unknown and the partial performances I don't know if we agree or disagree? Let me try again: 1. A performance AR between a recording the appropriate movements is the most accurate we can do. Symphony No. 40 in G minor is a performance of Sinfonie in g K. 550: I. Molto Allegro (2. Fassung) Symphony No. 40 in G minor is a performance of Sinfonie in g K. 550: II. Andante (2. Fassung) 2. The accuracy of a performance AR between a recording a super-work is depending on the parts linked to the super-work. Symphony No. 40 in G minor is a performance of Sinfonie in g K. 550 Which version? (Works structure isn't clearly defined yet, this super-work could contain 2 superworks or 8 movements. But the issue regarding accuracy remains.) I believe you are thinking as if only an AR to a super-work could be poorly defined. I believe that many ARs to movements are inaccurate. You ask me which version of the version? I ask you which version of the movements were really used in the currently existing movement ARs? Are you really sure that for each AR to a movement the full movement was really performed? Do you really think that the performer never missed a repeat? And how are we going to link to the correct movement versions all the recordings which already exist in MB? Our data is already inaccurate, and although the Work system should help improve things, the problem is never going to disappear completely. Links to works are often going to be fuzzy, because available information is fuzzy, and downloading music without booklets is not going to improve the situation. I believe the accuracy of the AR to movements is fallacious. I don't really think a performance of a Work should be considered less accurate than performances of each movements. If a performer performed a whole work as a whole, my first idea is to link to that work as a whole. But if you really think linking to each movement is an improvement, then this linking to a superwork means inaccurate should be stated very clearly, because I think other users will feel like me. I'd even consider appending use only if you can't find better or use only if you're not sure after each super-work title. -- Frederic Da Vitoria (davitof) Membre de l'April - « promouvoir et défendre le logiciel libre » - http://www.april.org ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
Hello, On 18/07/11 14:43, Andii Hughes wrote: On 18 July 2011 13:32, Kuno Woudtk...@frob.nl wrote: http://wiki.musicbrainz.org/Style/Track_and_release_titles ^ that is the entire guideline which applies to track and release titles. There is no Featured artists guideline for track and release titles. That's just the old release title guideline. What happens with the artist credit? http://wiki.musicbrainz.org/Style/Release briefly mentions how to enter the release artist, we don't have have a page like that for tracks. -- kuno / warp. ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On Mon, 2011-07-18 at 11:19 +0200, SwissChris wrote: There seem to be several unresolved questions here we should try to address separately: 1. How does as on cover (on track/release level) ideally translate into NGS structures? For an album by artist Foo with a track Quux featuring artist Bar: A) [track] Quux (feat. Bar) [Artist Credit] Foo (as we had pre-NGS) or B) [track] Quux [Artist Credit] Foo feat. Bar (using the NGS multiple artist feature, with the link phrase exactly as on cover) ? I have a preference for B. 2. How should the featured artists ideally be treated on the (normalized) recording level? Assuming there is consensus that feat.-information should not be part of the title field at recording level: A) [Artist Credit] Foo Bar (as by default) B) [Artist Credit] Foo feat. Bar (assuming feat. has a specific different meaning, not to be normalized to ) I would strongly argue in favour of B. 3. Do we want/need a delay on application of (part of) these ideal rules, once agreed upon, because of the ws/1 issue? I don't want to, but I understand that this could impact ws/1. I guess I'd ask how quickly a relatively straightforward change to the ws/1 could be made? 4. Do any or all of the above rules apply only to the exact wording feat. (or featuring, features) or would it also affect (how?) other link terms (in other languages) like with, en duo avec, con I would prefer we keep feat. as the link phrase as it's a very commonly used abbreviation. The guideline should IMO reflect the ideal regulations, with a warning box for not yet to be used features because of ws/1 compatibility issues, which could be removed (without further debate) when it's no longer used. +1 Paul ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] CSG: Classical superworks and performance of links
On Mon, 18 Jul 2011 15:02:02 +0200, Frederic Da Vitoria davito...@gmail.com wrote: 2011/7/18, symphonick symphon...@gmail.com: 1. A performance AR between a recording the appropriate movements is the most accurate we can do. Symphony No. 40 in G minor is a performance of Sinfonie in g K. 550: I. Molto Allegro (2. Fassung) Symphony No. 40 in G minor is a performance of Sinfonie in g K. 550: II. Andante (2. Fassung) 2. The accuracy of a performance AR between a recording a super-work is depending on the parts linked to the super-work. Symphony No. 40 in G minor is a performance of Sinfonie in g K. 550 Which version? (Works structure isn't clearly defined yet, this super-work could contain 2 superworks or 8 movements. But the issue regarding accuracy remains.) I believe you are thinking as if only an AR to a super-work could be poorly defined. I believe that many ARs to movements are inaccurate. No I don't, but the issue here was super-works. I've also been thinking about what to do with cases like recording of unknown piano version of work x. Let's say we have 3 works in the db: original work x for guitar, work x (piano version by foo) work x (piano version by bar). I suppose link to the original work is better than creating a new fuzzy piano version. You ask me which version of the version? I ask you which version of the movements were really used in the currently existing movement ARs? I suppose that in this case we need to create 3 works for every movement (+ a corresponding super-work): version 1, version 2 a fuzzy version. Or we define works as more of the concept movement I say that all versions are the same work? Are you really sure that for each AR to a movement the full movement was really performed? Do you really think that the performer never missed a repeat? I'd consider that a performance variation, still the same work. Links to works are often going to be fuzzy I agree. So what I'm trying to say is that it is - as always - preferable to be as specific as possible, if the data is available. if we know exactly what 4 parts of a super-work is performed, IMO it's more accurate to link directly to those than to a super-work containing more than 4 parts. If there's no different versions or added movements to consider (like in the OP), then I'd choose the most convenient option. No difference in accuracy what I can see. /symphonick ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On 07/18/2011 05:30 AM, Andii Hughes wrote: 1. If the cover says 'X (feat. Y)' by Z, then artist credit is Z, title is 'X (feat. Y)' 2. If the cover says 'X' by Z feat. Y then artist credit is Z + join-phrase ' feat. ' + Y and title is 'X' 3. If the cover says 'X (with new singing sensation Y)' by Z, then artist credit is Z, title is 'X (with new singing sensation Y)' 4. If the cover says 'X' by Z with his best friend Y, then artist credit is Z + join-phrase ' with his best friend ' + Y and title is 'X'. and so on... i.e. whatever is on the cover... +1 to all of this. Those of us (myself, for one) who want more normalized titles can have their tagger look at the recordings or the works instead of the tracklists. The one downside I can see to this is that the feat. artist will not have that release appear in their releases because they don’t have an artist credit on it (it will only appear in relationships). I don’t know if that’s enough of a problem for it to mean we need an exception guideline to the “as on the cover” concept. —Alex Mauer “hawke” ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-331: Add CD Baby Relationship Type
And we will just end up with ton of spam links that will clutter the database and web site. It doesn't necessarily need to clutter the site. When looking at a release there could be a single find this for sale button showing that info. ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
If we go this literally with as on cover for tracklists, and therefore normalize recordings to have all artists in the artist field, not in the title field, then... Can the server look at the recordings attached to a release's tracklist to see the artist credits of the recordings? If so, the server could theoretically list the release on the pages of artists which show up in the recordings but are missing in the tracklist. On Mon, Jul 18, 2011 at 1:02 PM, Alex Mauer ha...@hawkesnest.net wrote: On 07/18/2011 05:30 AM, Andii Hughes wrote: 1. If the cover says 'X (feat. Y)' by Z, then artist credit is Z, title is 'X (feat. Y)' 2. If the cover says 'X' by Z feat. Y then artist credit is Z + join-phrase ' feat. ' + Y and title is 'X' 3. If the cover says 'X (with new singing sensation Y)' by Z, then artist credit is Z, title is 'X (with new singing sensation Y)' 4. If the cover says 'X' by Z with his best friend Y, then artist credit is Z + join-phrase ' with his best friend ' + Y and title is 'X'. and so on... i.e. whatever is on the cover... +1 to all of this. Those of us (myself, for one) who want more normalized titles can have their tagger look at the recordings or the works instead of the tracklists. The one downside I can see to this is that the feat. artist will not have that release appear in their releases because they don’t have an artist credit on it (it will only appear in relationships). I don’t know if that’s enough of a problem for it to mean we need an exception guideline to the “as on the cover” concept. —Alex Mauer “hawke” ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
This convinces me that if we choose to keep track titles—which I think we probably will—then they should be as-printed, whereas recordings should be normalized. Given this, I revise my opinion accordingly: track titles should contain feat. ETI and recordings should use the feat. link phrase. Paul On Mon, 2011-07-18 at 13:51 +0200, Frederic Da Vitoria wrote: 2011/7/18, Kuno Woudt k...@frob.nl: Hello, On 18/07/11 12:14, Andii Hughes wrote: That's the whole point of starting with a (mostly) blank slate for parts which are largely new for NGS. To not impose new guidelines without going through the usual process. And something has to hold while new guidelines are being prepared. To say these interim guidelines are anything other than the previously approved ones is imposing new ones. If you think so, then so be it, we've imposed new guidelines. We've worked on the NGS guidelines for a year before the NGS release. During that year we've had many discussions on how to deal with these new entities. At various points throughout the process we've asked for feedback and help, through this mailinglist and on irc. If you're unhappy with the approach we've chosen, it's a bit late too complain about it now, there was more than enough time to give us that feedback before the NGS release. Like Andii, I would NOT use un-normalized data. I'd even be more extreme, I'd entirely remove things like feat which don't make much sense to me. From my point of view, an artist either performed or he did not, period. BUT - one frequent issue with edits before NGS was fights beween pro-normalization and pro-as-printed - one of the proposed benefits of NGS was allowing to enter data as printed as well as normalized. So that now I really don't know how I'd explain to a pro-as-printed: ok, we sold you NGS as solving all the issues, now NGS is here, it is much more clumsy to use than pre-NGS, there are almost twice as much fields, but the fields where you can enter data as printed are.. well, nowhere. I agree that entering data in 2 conflicting ways is going to be complicated. But if we don't, then I really fail to see what track titles and release titles are for. Let's do with recording titles and release group titles only, and we'll be perfectly normalized. I am quite serious here, let's remove track and release titles entirely, I won't complain, recording titles and release group titles are probably all I will ever use and data input will obviously be much simpler. And don't try to sell me a 2-level normalization system. I already had difficulties to memorize a few of the mysterious rules of the pre-NGS one-level system, I'd never manage to learn a 2-level system. ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On Mon, Jul 18, 2011 at 7:02 PM, Alex Mauer ha...@hawkesnest.net wrote: On 07/18/2011 05:30 AM, Andii Hughes wrote: 1. If the cover says 'X (feat. Y)' by Z, then artist credit is Z, title is 'X (feat. Y)' 2. If the cover says 'X' by Z feat. Y then artist credit is Z + join-phrase ' feat. ' + Y and title is 'X' 3. If the cover says 'X (with new singing sensation Y)' by Z, then artist credit is Z, title is 'X (with new singing sensation Y)' 4. If the cover says 'X' by Z with his best friend Y, then artist credit is Z + join-phrase ' with his best friend ' + Y and title is 'X'. and so on... i.e. whatever is on the cover... +1 to all of this. Sorry again, guys, but that's Taliban Speech. We're here in a database, and a database does build structures to store information from various sources in a comprehensive, usable form: artist information in artist field, track information in track fields, Extra Title Information in appropriate Comment or Annotation fields (wherever these may be found on the original source). If feat. should remain in the track title, as on the cover then why not keep everything else there as well just because it happens to follow directly after a track title? Why should we keep what in fact is Extra *Artist *Information in the track title, when stuff that actually is Extra *Title* Information (like: from West Side Story) should be moved to some other field? I have a disc in hand (Ute Lemper sings Kurt Weill) where a random track title reads 7. NANNA'S LIED - BRECHT/WEILL - Piano: Kai Rautenberg - BROOKHOUSE MUSIC INC. And this should all go in track title, because it's as on the cover? We have agreed (e.g. for capitalization guidelines) that you can't deduce artist intent from covers, since what's on a cover is mostly graphist's intent. Just because for obvious redundancy and aesthetic reasons a main artist is not generally repeated on every track, it's no longer as on the cover, when we add him? As said before all collaborations (be it feat., with, en duo avec) semantically, logically link two artists, never a track title to an artist. So track (feat. artist) on a cover is just a shortcut for track ([by artist] feat. artist). The NGS developers have done a great job on behalf of those that wanted as on cover presentation of the data (ok. They still haven't solved the problem of font size color, but I'm sure they're working on this flaw ;-) The NGS artist credit allows you to credit a featuring artist – in the Artist Credits – with the exact wording you find on the cover, with the one tiny exception that the Main Artist (flowing down from the release level to the track level, as Andii puts it) has to be included. And now, just because of this, you simply reject it? If this is the way it should be, I'd vote with Frederic to just forget about the two level system completely, which right now makes editing much more time-consuming and complicated. Chris/chabreyflint Those of us (myself, for one) who want more normalized titles can have their tagger look at the recordings or the works instead of the tracklists. The one downside I can see to this is that the feat. artist will not have that release appear in their releases because they don’t have an artist credit on it (it will only appear in relationships). I don’t know if that’s enough of a problem for it to mean we need an exception guideline to the “as on the cover” concept. —Alex Mauer “hawke” ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On 18/07/11 14:00, MeinDummy wrote: swisschris wrote: There seem to be several unresolved questions here we should try to address separately: 1. How does as on cover (on track/release level) *ideally* translate into NGS structures? For an album by artist Foo with a track Quux featuring artist Bar: A) [track] Quux (feat. Bar) [Artist Credit] Foo (as we had pre-NGS) or B) [track] Quux [Artist Credit] Foo feat. Bar (using the NGS multiple artist feature, with the link phrase exactly as on cover) ? B. 'As printed' does IMO not mean that additional artist info has to be part of the track title if the cover happens to have them next to each other. Otherwise we'd end up with A for all 'feat.' tracks on single artist releases and with B for almost all occurrences of the same track on VA compilations. You say that as if it's a bad thing! I mean, isn't that what happens. Nick Cave The Bad Seeds' Where the Wild Roses Grow barely mentions Kylie Minogue on the original album [1] and their best of [2], but on Ultimate Kylie, it's by Kylie Minogue Nick Cave The Bad Seeds [3]. Wasn't one of the goals of NGS to avoid having to normalise all of them, because now the recording links to them all? http://wiki.musicbrainz.org/Consistent_Original_Data is glorious history... Also, I think there's a subtle difference between a track/recording featuring a guest artist, and an artist with featured performer(s) [4], which tends to be muddied by converting them to artist credits. 1: http://www.discogs.com/viewimages?release=556821 2: http://www.discogs.com/viewimages?release=745099 3: http://www.discogs.com/viewimages?release=349240 4: http://www.bbc.co.uk/music/reviews/j2fd ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On Mon, 2011-07-18 at 20:30 +0100, Simon Austin wrote: You say that as if it's a bad thing! I mean, isn't that what happens. Nick Cave The Bad Seeds' Where the Wild Roses Grow barely mentions Kylie Minogue on the original album [1] and their best of [2], but on Ultimate Kylie, it's by Kylie Minogue Nick Cave The Bad Seeds [3]. Wasn't one of the goals of NGS to avoid having to normalise all of them, because now the recording links to them all? http://wiki.musicbrainz.org/Consistent_Original_Data is glorious history... Huh, I wonder why that page was marked as history? Although it could do with some wording updates, the principal is still valid, in particular for the two cases I can think of: * You cannot find any primary sources for the tracklisting, and are using the combination of several secondary sources (fan pages, Wikipedia, stores, etc.) to determine the tracklisting. * You are determining the recording title for a recording with name variations on different releases, where you can't determine a definite artist intent for a particular style. It seems a bit premature to just make the style principal go away like that. -- Calvin Walton calvin.wal...@kepstin.ca ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On Mon, Jul 18, 2011 at 9:30 PM, Simon Austin chi...@auzsoft.net wrote: On 18/07/11 14:00, MeinDummy wrote: swisschris wrote: There seem to be several unresolved questions here we should try to address separately: 1. How does as on cover (on track/release level) *ideally* translate into NGS structures? For an album by artist Foo with a track Quux featuring artist Bar: A) [track] Quux (feat. Bar) [Artist Credit] Foo (as we had pre-NGS) or B) [track] Quux [Artist Credit] Foo feat. Bar (using the NGS multiple artist feature, with the link phrase exactly as on cover) ? B. 'As printed' does IMO not mean that additional artist info has to be part of the track title if the cover happens to have them next to each other. Otherwise we'd end up with A for all 'feat.' tracks on single artist releases and with B for almost all occurrences of the same track on VA compilations. You say that as if it's a bad thing! I mean, isn't that what happens. Nick Cave The Bad Seeds' Where the Wild Roses Grow barely mentions Kylie Minogue on the original album [1] and their best of [2], but on Ultimate Kylie, it's by Kylie Minogue Nick Cave The Bad Seeds [3]. Wasn't one of the goals of NGS to avoid having to normalise all of them, because now the recording links to them all? http://wiki.musicbrainz.org/Consistent_Original_Data is glorious history... I was going to bring exactly the same example of Where the Wild Roses grow. You forgot to mention the Single which credits Nick Cave and The Bad Seeds *+* Kylie Minogue [1] and all the compilations which have (wrong or right) feat. Kylie Minogue (as recordings!) on MB [2]. Fact is that NGS artist credit allows to catch all these differences quite precisely as on cover. I'd also use this example to show that feat is not always so specific in meaning as to have it kept by all means on recording level. On the other side this example shows that you probably can't trust all the feat. on MB tracks to really be feat. on the cover, since the previous guideline asked to use this specific term for other link phrases as well… [1] http://www.discogs.com/viewimages?release=833051 [2] http://musicbrainz.org/artist/172e1f1a-504d-4488-b053-6344ba63e6d0/recordings?page=25 Also, I think there's a subtle difference between a track/recording featuring a guest artist, and an artist with featured performer(s) [4], which tends to be muddied by converting them to artist credits. This last example is a temporary issue and should be fixed (catching the actual Artist Credits AFAIK) when ws/1 clients will change to ws/2 Chris/chabreyflint 1: http://www.discogs.com/viewimages?release=556821 2: http://www.discogs.com/viewimages?release=745099 3: http://www.discogs.com/viewimages?release=349240 4: http://www.bbc.co.uk/music/reviews/j2fd ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On 18 July 2011 21:02, SwissChris swissch...@gmail.com wrote: snip.. I was going to bring exactly the same example of Where the Wild Roses grow. You forgot to mention the Single which credits Nick Cave and The Bad Seeds + Kylie Minogue [1] and all the compilations which have (wrong or right) feat. Kylie Minogue (as recordings!) on MB [2]. Fact is that NGS artist credit allows to catch all these differences quite precisely as on cover. I'd also use this example to show that feat is not always so specific in meaning as to have it kept by all means on recording level. On the other side this example shows that you probably can't trust all the feat. on MB tracks to really be feat. on the cover, since the previous guideline asked to use this specific term for other link phrases as well… [1] http://www.discogs.com/viewimages?release=833051 [2] http://musicbrainz.org/artist/172e1f1a-504d-4488-b053-6344ba63e6d0/recordings?page=25 Most of those probably should be merged, which means one still needs to be picked for the recording. Personally, I'd tend to trust the single cover as it's the case where the most focus has been on that particular track, whereas with an album it's one of many. Also, I think there's a subtle difference between a track/recording featuring a guest artist, and an artist with featured performer(s) [4], which tends to be muddied by converting them to artist credits. This last example is a temporary issue and should be fixed (catching the actual Artist Credits AFAIK) when ws/1 clients will change to ws/2 It needs splitting on MB, but yes the BBC still using ws/1 is one of the things I was worried about. You'll get exactly that situation where there's one single artist composed of the primary and featured artists. Chris/chabreyflint 1: http://www.discogs.com/viewimages?release=556821 2: http://www.discogs.com/viewimages?release=745099 3: http://www.discogs.com/viewimages?release=349240 4: http://www.bbc.co.uk/music/reviews/j2fd ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style -- Andii :-) ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On 18 July 2011 14:00, Kuno Woudt k...@frob.nl wrote: Hello, On 18/07/11 14:43, Andii Hughes wrote: On 18 July 2011 13:32, Kuno Woudtk...@frob.nl wrote: http://wiki.musicbrainz.org/Style/Track_and_release_titles ^ that is the entire guideline which applies to track and release titles. There is no Featured artists guideline for track and release titles. That's just the old release title guideline. What happens with the artist credit? http://wiki.musicbrainz.org/Style/Release briefly mentions how to enter the release artist, we don't have have a page like that for tracks. -- kuno / warp. ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style Ok, let's take a step back then. If the current guideline only applies to recordings, should I alter this proposal to also just concern recordings and release groups (i.e. the bit about moving the feat. part to the artist credit)? I think it should have been made much clearer that there weren't any rules for track or release titles, because it certainly wasn't clear from the discussions I've followed. Also, I can understand the choice between track and recording titles for unnormalised/normalised data, but is the same thing supposed to apply for releases and release groups? -- Andii :-) ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On Mon, Jul 18, 2011 at 11:13 PM, Andii Hughes gnu_and...@member.fsf.org wrote: On 18 July 2011 14:00, Kuno Woudt k...@frob.nl wrote: Hello, On 18/07/11 14:43, Andii Hughes wrote: On 18 July 2011 13:32, Kuno Woudtk...@frob.nl wrote: http://wiki.musicbrainz.org/Style/Track_and_release_titles ^ that is the entire guideline which applies to track and release titles. There is no Featured artists guideline for track and release titles. That's just the old release title guideline. What happens with the artist credit? http://wiki.musicbrainz.org/Style/Release briefly mentions how to enter the release artist, we don't have have a page like that for tracks. -- kuno / warp. ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style Ok, let's take a step back then. If the current guideline only applies to recordings, should I alter this proposal to also just concern recordings and release groups (i.e. the bit about moving the feat. part to the artist credit)? I think that would avoid any vetoes and allow us to go forward. We can argue about tracks in another proposal :) Of course, some of the questions still apply: is con, that translates as with but is sometimes used as feat., to be changed to with? to feat.? :) I think it should have been made much clearer that there weren't any rules for track or release titles, because it certainly wasn't clear from the discussions I've followed. Also, I can understand the choice between track and recording titles for unnormalised/normalised data, but is the same thing supposed to apply for releases and release groups? At least in theory, yes. -- Andii :-) ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style -- Nicolás Tamargo de Eguren ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On 18/07/11 21:02, SwissChris wrote: Also, I think there's a subtle difference between a track/recording featuring a guest artist, and an artist with featured performer(s) [4], which tends to be muddied by converting them to artist credits. This last example is a temporary issue and should be fixed (catching the actual Artist Credits AFAIK) when ws/1 clients will change to ws/2 4: http://www.bbc.co.uk/music/reviews/j2fd To clarify... The Zombies started in the 60s and split up at the end of that decade. The 2011 album by The Zombies featuring Colin Blunstone Rod Argent highlights the two original members who reformed the band. They're not guests, nor collaborating. - Si ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
Hello, On 18/07/11 21:45, Calvin Walton wrote: On Mon, 2011-07-18 at 20:30 +0100, Simon Austin wrote: You say that as if it's a bad thing! I mean, isn't that what happens. Nick Cave The Bad Seeds' Where the Wild Roses Grow barely mentions Kylie Minogue on the original album [1] and their best of [2], but on Ultimate Kylie, it's by Kylie Minogue Nick Cave The Bad Seeds [3]. Wasn't one of the goals of NGS to avoid having to normalise all of them, because now the recording links to them all? http://wiki.musicbrainz.org/Consistent_Original_Data is glorious history... Huh, I wonder why that page was marked as history? That guideline was moved to the Style Principle page: http://wiki.musicbrainz.org/Style/Principle If no definite proof can be found for the correct spelling/punctuation, the most common version should be used. -- kuno / warp. ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] CSG: Classical superworks and performance of links
2011/7/18 symphonick symphon...@gmail.com On Mon, 18 Jul 2011 15:02:02 +0200, Frederic Da Vitoria davito...@gmail.com wrote: 2011/7/18, symphonick symphon...@gmail.com: 1. A performance AR between a recording the appropriate movements is the most accurate we can do. Symphony No. 40 in G minor is a performance of Sinfonie in g K. 550: I. Molto Allegro (2. Fassung) Symphony No. 40 in G minor is a performance of Sinfonie in g K. 550: II. Andante (2. Fassung) 2. The accuracy of a performance AR between a recording a super-work is depending on the parts linked to the super-work. Symphony No. 40 in G minor is a performance of Sinfonie in g K. 550 Which version? (Works structure isn't clearly defined yet, this super-work could contain 2 superworks or 8 movements. But the issue regarding accuracy remains.) I believe you are thinking as if only an AR to a super-work could be poorly defined. I believe that many ARs to movements are inaccurate. No I don't, but the issue here was super-works. I've also been thinking about what to do with cases like recording of unknown piano version of work x. Let's say we have 3 works in the db: original work x for guitar, work x (piano version by foo) work x (piano version by bar). I suppose link to the original work is better than creating a new fuzzy piano version. You ask me which version of the version? I ask you which version of the movements were really used in the currently existing movement ARs? I suppose that in this case we need to create 3 works for every movement (+ a corresponding super-work): version 1, version 2 a fuzzy version. Or we define works as more of the concept movement I say that all versions are the same work? Are you really sure that for each AR to a movement the full movement was really performed? Do you really think that the performer never missed a repeat? I'd consider that a performance variation, still the same work. Links to works are often going to be fuzzy I agree. So what I'm trying to say is that it is - as always - preferable to be as specific as possible, if the data is available. if we know exactly what 4 parts of a super-work is performed, IMO it's more accurate to link directly to those than to a super-work containing more than 4 parts. If there's no different versions or added movements to consider (like in the OP), then I'd choose the most convenient option. No difference in accuracy what I can see. And create 32 ARs for a global performance of the Goldberg Variations? It seems a little overkill to me, but if you feel it is really important, I won't fight it. It won't happen often. I still don't like saying that an AR to a super-work means fuzzy, I'd rather have an explicit fuzzy attribute, but if others agree your solution is better... -- Frederic Da Vitoria (davitof) Membre de l'April - « promouvoir et défendre le logiciel libre » - http://www.april.org ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
Restricting the guideline to recordings release groups would certainly make things go smoother. For the ws/1 issue i'd see something like we have here: http://wiki.musicbrainz.org/Theatre_Style Chris/chabreyflint On Mon, Jul 18, 2011 at 10:13 PM, Andii Hughes gnu_and...@member.fsf.orgwrote: On 18 July 2011 14:00, Kuno Woudt k...@frob.nl wrote: Hello, On 18/07/11 14:43, Andii Hughes wrote: On 18 July 2011 13:32, Kuno Woudtk...@frob.nl wrote: http://wiki.musicbrainz.org/Style/Track_and_release_titles ^ that is the entire guideline which applies to track and release titles. There is no Featured artists guideline for track and release titles. That's just the old release title guideline. What happens with the artist credit? http://wiki.musicbrainz.org/Style/Release briefly mentions how to enter the release artist, we don't have have a page like that for tracks. -- kuno / warp. ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style Ok, let's take a step back then. If the current guideline only applies to recordings, should I alter this proposal to also just concern recordings and release groups (i.e. the bit about moving the feat. part to the artist credit)? I think it should have been made much clearer that there weren't any rules for track or release titles, because it certainly wasn't clear from the discussions I've followed. Also, I can understand the choice between track and recording titles for unnormalised/normalised data, but is the same thing supposed to apply for releases and release groups? -- Andii :-) ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] CSG: Classical superworks and performance of links
On Mon, 18 Jul 2011 22:23:14 +0200, Frederic Da Vitoria davito...@gmail.com wrote: I agree. So what I'm trying to say is that it is - as always - preferable to be as specific as possible, if the data is available. if we know exactly what 4 parts of a super-work is performed, IMO it's more accurate to link directly to those than to a super-work containing more than 4 parts. If there's no different versions or added movements to consider (like in the OP), then I'd choose the most convenient option. No difference in accuracy what I can see. And create 32 ARs for a global performance of the Goldberg Variations? It seems a little overkill to me, but if you feel it is really important, I won't fight it. It won't happen often. I still don't like saying that an AR to a super-work means fuzzy, I'd rather have an explicit fuzzy attribute, but if others agree your solution is better... I don't know if it's important or not, maybe the inaccurate cases are rare, so we don't have to worry about it. I'm hoping we can figure out what we want from works; what level of detail can we manage? If we forget my add a fuzzy work suggestion, and try your fuzzy performance, would it be something like this: There are 2 super-works in the db: Sinfonie in g K. 550 (1. Fassung) Sinfonie in g K. 550 (2. Fassung) (assuming for now that this is the default) and we have a performance where we don't know the version: Symphony No. 40 in G minor is a (assumed) performance of Sinfonie in g K. 550 (2. Fassung) (someone can find a better term) /symphonick ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On Mon, Jul 18, 2011 at 10:02 AM, Alex Mauer ha...@hawkesnest.net wrote: On 07/18/2011 05:30 AM, Andii Hughes wrote: 1. If the cover says 'X (feat. Y)' by Z, then artist credit is Z, title is 'X (feat. Y)' 2. If the cover says 'X' by Z feat. Y then artist credit is Z + join-phrase ' feat. ' + Y and title is 'X' 3. If the cover says 'X (with new singing sensation Y)' by Z, then artist credit is Z, title is 'X (with new singing sensation Y)' 4. If the cover says 'X' by Z with his best friend Y, then artist credit is Z + join-phrase ' with his best friend ' + Y and title is 'X'. and so on... i.e. whatever is on the cover... +1 to all of this. +1 again. The one downside I can see to this is that the feat. artist will not have that release appear in their releases because they don’t have an artist credit on it (it will only appear in relationships). I don’t know if that’s enough of a problem for it to mean we need an exception guideline to the “as on the cover” concept. Seems to me like this credit *should* appear in relationships for the guest artist, because that's (usually) the most accurate description of how involved they were with the track. For that matter, has anybody suggested NOT listing the featured guest as an artist, but instead adding a {featured} option to the performer AR and adding the guest there? --Torc. ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On 18 July 2011 23:50, Ryan Torchia anarchyr...@gmail.com wrote: On Mon, Jul 18, 2011 at 10:02 AM, Alex Mauer ha...@hawkesnest.net wrote: On 07/18/2011 05:30 AM, Andii Hughes wrote: 1. If the cover says 'X (feat. Y)' by Z, then artist credit is Z, title is 'X (feat. Y)' 2. If the cover says 'X' by Z feat. Y then artist credit is Z + join-phrase ' feat. ' + Y and title is 'X' 3. If the cover says 'X (with new singing sensation Y)' by Z, then artist credit is Z, title is 'X (with new singing sensation Y)' 4. If the cover says 'X' by Z with his best friend Y, then artist credit is Z + join-phrase ' with his best friend ' + Y and title is 'X'. and so on... i.e. whatever is on the cover... +1 to all of this. +1 again. The one downside I can see to this is that the feat. artist will not have that release appear in their releases because they don’t have an artist credit on it (it will only appear in relationships). I don’t know if that’s enough of a problem for it to mean we need an exception guideline to the “as on the cover” concept. Seems to me like this credit *should* appear in relationships for the guest artist, because that's (usually) the most accurate description of how involved they were with the track. I think it should and that would be the benefit of moving to artist rather than track accreditation for recordings. For that matter, has anybody suggested NOT listing the featured guest as an artist, but instead adding a {featured} option to the performer AR and adding the guest there? That's part of the current and proposed guideline: 'In both cases: Add Relationships of the appropriate Relationship Class (usually Performance) to link to the featured artist(s') entries in MusicBrainz.' There's a 'guest' tick box for this. I've added many such ARs in the past. --Torc. ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style -- Andii :-) ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On Tue, Jul 19, 2011 at 1:53 AM, Andii Hughes gnu_and...@member.fsf.org wrote: On 18 July 2011 23:50, Ryan Torchia anarchyr...@gmail.com wrote: On Mon, Jul 18, 2011 at 10:02 AM, Alex Mauer ha...@hawkesnest.net wrote: On 07/18/2011 05:30 AM, Andii Hughes wrote: 1. If the cover says 'X (feat. Y)' by Z, then artist credit is Z, title is 'X (feat. Y)' 2. If the cover says 'X' by Z feat. Y then artist credit is Z + join-phrase ' feat. ' + Y and title is 'X' 3. If the cover says 'X (with new singing sensation Y)' by Z, then artist credit is Z, title is 'X (with new singing sensation Y)' 4. If the cover says 'X' by Z with his best friend Y, then artist credit is Z + join-phrase ' with his best friend ' + Y and title is 'X'. and so on... i.e. whatever is on the cover... +1 to all of this. +1 again. The one downside I can see to this is that the feat. artist will not have that release appear in their releases because they don’t have an artist credit on it (it will only appear in relationships). I don’t know if that’s enough of a problem for it to mean we need an exception guideline to the “as on the cover” concept. Seems to me like this credit *should* appear in relationships for the guest artist, because that's (usually) the most accurate description of how involved they were with the track. I think it should and that would be the benefit of moving to artist rather than track accreditation for recordings. For that matter, has anybody suggested NOT listing the featured guest as an artist, but instead adding a {featured} option to the performer AR and adding the guest there? That's part of the current and proposed guideline: 'In both cases: Add Relationships of the appropriate Relationship Class (usually Performance) to link to the featured artist(s') entries in MusicBrainz.' There's a 'guest' tick box for this. I've added many such ARs in the past. Not *exactly* the same, though, even if fairly similar: there are some cases where the liner credits say and guest guitar X without being considered a feat. But quite close, true that. --Torc. ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style -- Andii :-) ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style -- Nicolás Tamargo de Eguren ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
On 18/07/11 12:07, SwissChris wrote: When cover says Track (feat. Artist) should this (follow the cover) translate into Track (feat. artist) (on track field) [by] Main Artist (on AC field) or (as I think is the only reasonable way) as Track (on track field) [by] Main Artist feat. featured artist (on AC field). if we don't want ws/1 compatibility i would favour to move feat. also on tracklvl to the artist credits, away from the title field. Am 18.07.2011, 12:31 Uhr, schrieb Kuno Woudt k...@frob.nl: Here are two examples (warning, large images :). Track 14 on Mysteryland 2010 is listed on the cover as: track title: Born Again artist credits: RICKY L FEAT. M:CK http://www.frob.nl/mb/2011/mysteryland.black.2010.jpg Track 1 on BoA is listed on the cover as: track title: I Did It For Love (featuring Sean Garret) artist credits: BoA http://www.frob.nl/mb/2011/smusa01.jpg these are some nice examples! 1. compilation where the artist is credited on every track, and if there is a feat. artist it's next to the main artist. 2. a single artist cd, where there is no column for the artist (because of obvious reasons). however the feat. artist can't be appended to the artist here so it is written in parenthesis AND with smaller fontsize thus making it clearly separated from the track title. so beside that i think that we should credit/link featured artists properly also on track-lvl i do think that in most of these cases the feat. should not be part of the tracktitle in most occasions. if the feat. artist is not in parentheses and written in the same size/style as the rest of the track i would keep it in the tracktitle. -- lorenz pressler PGP 0x92E9551A ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
[mb-style] RFV-326: Add Artist Type Other
Hello, 7 days have passed, since there was no veto I submit the RFV. The proposal is to add a new Artist Type Other. (Minor) changes since RFC: All attributes / link phrases etc. should be the same as with type Unknown. Wiki page: http://wiki.musicbrainz.org/Proposal:Artist_Type_Other Expiration date: Wed, 20 Jul 2011 23:00 UTC Johannes ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
2011/7/18 Nicolás Tamargo de Eguren reosare...@gmail.com: On Mon, Jul 18, 2011 at 11:13 PM, Andii Hughes snip... Ok, let's take a step back then. If the current guideline only applies to recordings, should I alter this proposal to also just concern recordings and release groups (i.e. the bit about moving the feat. part to the artist credit)? I think that would avoid any vetoes and allow us to go forward. We can argue about tracks in another proposal :) Done; see the revised version at http://wiki.musicbrainz.org/Proposal:Featured_Artists Of course, some of the questions still apply: is con, that translates as with but is sometimes used as feat., to be changed to with? to feat.? :) I don't really feel qualified to judge on these (the only ones I've ever seen are 'with' and 'feat.') and the previous guideline didn't cover them either. I'd say we need a separate guideline that maps them to either 'feat.' or ' ' (collaboration) at recording level. For the featured artist proposal, the join terms are mandated as 'feat.' then ', ' and ' ' for multiple featured artists. FWIW, 'with' sounds more like ' ' than 'feat.' to me. -- Andii :-) ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
Re: [mb-style] RFC-327: Featured Artists
Until now it was clear: If an artist (featured or guest or collaboration) was mentioned on the (back) cover he got artist credit (and the appropriate AR), if not he got only AR (where you could choose nearly everything from guest lead pedal guitar to additional soprano backing vocal). Since multiple artists were not supported we opted for a workaround solution that had Artist Credit for featured and similar artists, when on cover, appended in brackets in the title field. Now that MB supports multiple artists, allowing every link phrase one can think of, what speaks against moving them to Artist Credit where they belong (once the ws/1 issue solved)? For Where the Wild Roses grow we'd have - on the original Murder Ballads Album: AC Nick Cave; Kylie Minogue (not on the cover) added by appropriate AR - on the Single: joint AC Nick Kylie (or: Nick + Kylie) - on the Kylie Best of: AC to Kylie, AR for Nick (not on the cover) - on all the compilations where this track appears with both names: AC to both, linked by whatever link phrase (feat., with, duo with, +) that is on the cover. What we still have to agree on is what link phrase to use on a normalized recording level when we have various versions to choose from like above or when the link phrase on the cover is not in English… 2011/7/19 Nicolás Tamargo de Eguren reosare...@gmail.com On Tue, Jul 19, 2011 at 1:53 AM, Andii Hughes gnu_and...@member.fsf.org wrote: On 18 July 2011 23:50, Ryan Torchia anarchyr...@gmail.com wrote: On Mon, Jul 18, 2011 at 10:02 AM, Alex Mauer ha...@hawkesnest.net wrote: On 07/18/2011 05:30 AM, Andii Hughes wrote: 1. If the cover says 'X (feat. Y)' by Z, then artist credit is Z, title is 'X (feat. Y)' 2. If the cover says 'X' by Z feat. Y then artist credit is Z + join-phrase ' feat. ' + Y and title is 'X' 3. If the cover says 'X (with new singing sensation Y)' by Z, then artist credit is Z, title is 'X (with new singing sensation Y)' 4. If the cover says 'X' by Z with his best friend Y, then artist credit is Z + join-phrase ' with his best friend ' + Y and title is 'X'. and so on... i.e. whatever is on the cover... +1 to all of this. +1 again. The one downside I can see to this is that the feat. artist will not have that release appear in their releases because they don’t have an artist credit on it (it will only appear in relationships). I don’t know if that’s enough of a problem for it to mean we need an exception guideline to the “as on the cover” concept. Seems to me like this credit *should* appear in relationships for the guest artist, because that's (usually) the most accurate description of how involved they were with the track. I think it should and that would be the benefit of moving to artist rather than track accreditation for recordings. For that matter, has anybody suggested NOT listing the featured guest as an artist, but instead adding a {featured} option to the performer AR and adding the guest there? That's part of the current and proposed guideline: 'In both cases: Add Relationships of the appropriate Relationship Class (usually Performance) to link to the featured artist(s') entries in MusicBrainz.' There's a 'guest' tick box for this. I've added many such ARs in the past. Not *exactly* the same, though, even if fairly similar: there are some cases where the liner credits say and guest guitar X without being considered a feat. But quite close, true that. --Torc. ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style -- Andii :-) ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style -- Nicolás Tamargo de Eguren ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style ___ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style