Release Groups guideline
|
View:
New views
12 Messages
—
Rating Filter:
Alert me
|
 |
Release Groups guideline
by Kuno Woudt-2
::
Rate this Message:
Hello, I've seen the recent discussion about the bootleg stuff in the RG proposal, but I haven't been following the RG proposal from the start, so I ask this question: Are there any other issues which need to be taken care before we can RFV it? If so, please point me at the previous discussions, I want to keep this thing moving towards official as fast as is possible here on mb-style :) As far as I know, we don't currently have an IdeaChampion for the ReleaseGroups. Is anyone interested in picking that up? I think the first task would be to change the wording of the bootleg stuff, as apart from BrianFreud there mostly seems to be consensus on what to do there. (If you do pick this up, as always please make a copy as wiki.musicbrainz.org/User:yourname/ReleaseGroup, thanks). -- kuno / warp. _______________________________________________ Musicbrainz-style mailing list Musicbrainz-style@... http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style |
|
|
Re: Release Groups guideline
by Brian Schweitzer
::
Rate this Message:
On Mon, Aug 3, 2009 at 4:41 PM, Kuno Woudt <kuno@...> wrote:
A line by line glance before bed, a few issues, apart from the bootleg question, though I'm sure there's still issues here I've forgotten :P "Each disc in a multi-disc release at MusicBrainz is a separate release (named according to disc number style), but all discs should be part of the same release group. " The terminology here is correct for now, but will be incorrect come NGS. Actually, this entire line becomes redundant under NGS, as the whole multi-disc problem should likely move to the release and medium pages... "Multi-disc releases" vs "Box sets" sections - "A special case of the above; each disc in a box set would normally be part of the same release group. " - isn't the box set line redundant, esp as it doesn't define anything about when the discs of a box set would *not* be in the same release group? (or at least, not until the When Not to Group Releases section) "Promotional and pirated versions" vs "Promotional and bootleg versions" - I know navap doesn't like the word "pirated" being used here; I switched the wordage from "bootleg" to avoid the "what is intended" confusion problem mentioned in the bootleg RG edit. ("bootleg" here meaning "pirated copies of official releases", not "bootleg" in the MusicBrainz standard general sense of pirated copies, 'real' bootlegs, and homebrews all being "bootleg" for release status.) In IRC conversation, however, neither of us could figure a good wording here that would keep "bootleg" in there somehow, clarify that it only means pirated copies, not ALL bootlegs, and yet not overcomplexify the wording. "Weezer's "Weezer (Red Album)" has ten editions in the database, some releases from different countries, some deluxe editions, and one transliteration." - this wording will become out of date the moment someone adds a second transliteration... "Mozart's 9 volume, 170 disc "Wolfgang Amadeus Mozart: Complete Works" box set" example - this RG actually, in of itself, is problematic... It's grouping 9 different box sets, each individually released before they were released as a box set. This example (and RG) actually seems pretty much in direct conflict with the What Not to Group line "A Series of different volumes that may have been released over time.", given that while yes, the box was released all at once, the releases in it are titled as they were in the volume box (and actually in the big box as well), with individual Volume identification, not identification as a whole box without separate volumes. Either http://musicbrainz.org/release-group/314eb0c2-4998-3a8b-a2b7-03ee4e8338b1.html , or one of the Brilliant Classics "Historic Russian Archives" sets would seem to make a more correct and clear example. (Navap, I'd suggest that the edit to merge all the separate volumes of the box set maybe should be reverted, if the guideline stays written as it is now...) "For Box Sets or Multi-Disc Releases where each disc has been released separately. If these individual discs have not been duplicated due to the rules outlined at "What Defines A Unique Release", they should not be grouped together as per the normal box set rules noted above." is a little awkward to parse; something like: "For Box Sets or Multi-Disc Releases where each disc has also been released individually, the individually released releases should not be grouped together. Instead, the releases should be duplicated for the box set listing, according to the guidelines outlined at "What Defines A Unique Release", and grouped according to the normal box set rules." would be clearer. "If you are unsure about other cases, please feel free to come on IRC, use the mailing list, or add a note to the discussion page. " - is this not true for to any guideline (and thus redundant) - plus, the discussion page or IRC is not, imho, the place people should be going to for style discussions for the inevitable cat corner cases... "In the case of re-releases with different titles, the principle of Consistent Original Data" - Given two releases of the same album, with entirely different titles, I'm not sure how COD is applicable; if our intent is that the original release's title has greater predominance, we should just say that. "The types of a release group are the same as those possible for Release Type and should normally be the same as all of the releases in that group." - This becomes incorrect under NGS, as releases having a type at all is a concept of the current schema, not the NGS schema (under which RGs have a type, releases don't). "...release-specific Extra Title Information added due to Disc Number Style." - this too becomes outdated with NGS, as disc numbers no longer will be in release titles, but will be a medium-specific field. "It is OK for a bonus disc EP, remix, live disc etc to be inside the same "Album" release group as the accompanying album disc. " - See above for how this breaks under NGS; there is no release "type", so all of these would have their type wiped away into whatever the RG type is. However, I cannot think of the correct case, per the RG guidelines, where any of these would be correct to be grouped with an album, anyhow - so is this line simply redundant? Missing situations: * There is no discussion of when to/not to group soundtracks. * While there is a mention of 2 different releases of the same musical+cast, I'd suggest that there should be an example specificly stating that "My Fair Lady (Original London Cast)" and "My Fair Lady (Original Broadway Cast)" should *not* be merged. * Nothing in the guideline about audiobooks/spoken word. The J K Rowling listing, at the moment, imho is 99% useless in terms of functionality - it looks more like it's being used as a pseduo-artist-release "author of book" AR, not a functional grouping of quite distinct releases. (And grouping by book breaks the moment you have a 2-in-1 audiobook, anyhow...) * Not grouping classical / opera / theater / etc by work, but by recording? (ie, don't make one giant "Mozart's Requiem K 626" RG for every single release of the Requiem, no matter by whom...) Brian _______________________________________________ Musicbrainz-style mailing list Musicbrainz-style@... http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style |
|
|
Re: Release Groups guideline
by Brian Schweitzer
::
Rate this Message:
I'm reading through some other RG merge edits, and lest anyone
think I'm imagining some of these "needs 'do not merge' guidelines" as a possible thing someone might try to
do... * Not grouping classical / opera / theater / etc by work, but by recording? (ie, don't make one giant "Mozart's Requiem K 626" RG for every single release of the Requiem, no matter by whom...) http://musicbrainz.org/show/edit/?editid=10678956 "BTW, what do you think about classical releases. If the different recordings of the same classical piece of word (e.g. an opera) are properly stored in MB (by the use of release and medium from the NGS), would it not make sense to put all of them into the same group? For example, you would have one group for the opera "Les contes d'Hoffmann" by Offenbach containing all its versions." ...shudder. :) And yet, that logic almost holds together, if you accept the logic for grouping audiobooks by book, bootlegs by concert, etc, this almost holds together - yet the concept has nothing to do with grouping distinct releases (and here, and in the next, both have any reasonable rationale for doing it killed off by the introduction of Works in NGS...) * While there is a mention of 2 different releases of the same musical+cast, I'd suggest that there should be an example specificly stating that "My Fair Lady (Original London Cast)" and "My Fair Lady (Original Broadway Cast)" should *not* be merged. And, based on this comment in the RG discussion page, also exceptions for remixes, re-recordings, etc. (The comment is a direct response to a question about needing a "do not merge different casts/recordings of musicals" line. It again is an example of someone taking the "group by some conceptual commonality" idea and combining RGs based not on some common semantic meaning - exactly what happens when, as I said in the bootleg discussion, we merge things simply because someone sees some commonality, and enough people, on a one-by-one basis, can be persuaded to vote for it. http://wiki.musicbrainz.org/Talk:Release_Group "I think re-recorded releases should be in the same group as original one, as they still keep the same concept. Same goes to remixes. I don't see why Prodoc want to keep them all separated. Remixed albums still bears "remix" mark in its release type, and it is also a part of the same work, of the same concept. So they should be kept in original's RG, similar to how navap pointed on Demos section on this page (I tend to agree with that). Once again, if we keep each variation of the same work as a separate RG, we lose all the advantage of such implementation. Here are some examples I worked with: RG with re-recording (edit) and RG with remix (edit) - these are all the same work, these are all in the same release group. --Atedos 20:17, 27 May 2009 (UTC) " ...and then you have audiobooks, where somehow, the edit with all debate got cancelled after a 0:2 ( http://musicbrainz.org/show/edit/?editid=10678956 ), but the rest here seem to all be passing as if merging all possible unrelated release groups into one "basis book" group is an undisputed guideline: http://musicbrainz.org/mod/search/results.html?minid=&artist_type=3&mod_type=67&moderator_type=0&moderator_id=95678&automod=&voter_id=95678&artist_id=59157&orderby=asc&maxid=&isreset=0&voter_type=0&offset=275 The main problem, though, is that it seems like we're getting a ton of "same something" RG merges, all without style list or guideline decision, setting precidents based on the opinion of whoever happens to (quickly, frequently, due to all the 3:0's) see the edit. In 3 months, we've had nearly sixty *thousand* RG merge edits. (59,854 as of now). Now, a lot of those have been modbot, but afaik, while we can filter FreeDB, there's no way to filter out Modbot, and all these quite debatable RG merges are getting flooded out by all the simple "merging 2 versions of an album" edits. While it's really simple to create RG merge edits, it's a pita to revert them, and a lot of these "same something" reasons are quite debatable. The sooner we can get an official guideline that addresses things in a reasonable manner, and which has clear "when not to merge" that cover the non-album cases (esp bootlegs, soundtracks, musicals, classical, opera, etc.), the better. Brian _______________________________________________ Musicbrainz-style mailing list Musicbrainz-style@... http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style |
|
|
Re: Release Groups guideline
by gnu_andrew
::
Rate this Message:
2009/8/4 Brian Schweitzer <brian.brianschweitzer@...>:
> A line by line glance before bed, a few issues, apart from the bootleg > question, though I'm sure there's still issues here I've forgotten :P > > "Each disc in a multi-disc release at MusicBrainz is a separate release > (named according to disc number style), but all discs should be part of the > same release group. " > > The terminology here is correct for now, but will be incorrect come NGS. > Actually, this entire line becomes redundant under NGS, as the whole > multi-disc problem should likely move to the release and medium pages... > Given that most guidelines will need updating when NGS is released, this should probably wait until then. Adding wording that talks about something that applies to a release that won't happen for at least two to three months is just going to confuse people. >... > "For Box Sets or Multi-Disc Releases where each disc has been released > separately. If these individual discs have not been duplicated due to the > rules outlined at "What Defines A Unique Release", they should not be > grouped together as per the normal box set rules noted above." is a little > awkward to parse; something like: > "For Box Sets or Multi-Disc Releases where each disc has also been released > individually, the individually released releases should not be grouped > together. Instead, the releases should be duplicated for the box set > listing, according to the guidelines outlined at "What Defines A Unique > Release", and grouped according to the normal box set rules." > would be clearer. > The second version reads worse to me. Firstly, the use of 'individually' twice in succession reads badly, and the rewording of the second part of the sentence makes it sound like the user should go ahead and duplicate in all cases, rather than _only_ under the cases outlined by 'What Defines a Unique Release'. How about something like: "For Box Sets or Multi-Disc Releases where each disc has also been released separately, the releases should not be grouped together. Instead, those releases which qualify as unique according to the guidelines outlined at "What Defines A Unique Release", should be duplicated and grouped according to the normal box set rules." > Brian Cheers, -- Andrew :-) Free Java Software Engineer Red Hat, Inc. (http://www.redhat.com) Support Free Java! Contribute to GNU Classpath and the OpenJDK http://www.gnu.org/software/classpath http://openjdk.java.net PGP Key: 94EFD9D8 (http://subkeys.pgp.net) Fingerprint: F8EF F1EA 401E 2E60 15FA 7927 142C 2591 94EF D9D8 _______________________________________________ Musicbrainz-style mailing list Musicbrainz-style@... http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style |
|
|
Re: Release Groups guideline
by Pavel Fedyakov
::
Rate this Message:
Brian Schweitzer wrote:
> I'm reading through some other RG merge edits, and lest anyone think > I'm imagining some of these "needs 'do not merge' guidelines" as a > possible thing someone might try to do... > > > * Not grouping classical / opera / theater / etc by work, but by > recording? (ie, don't make one giant "Mozart's Requiem K 626" RG > for every single release of the Requiem, no matter by whom...) > > > http://musicbrainz.org/show/edit/?editid=10678956 > "BTW, what do you think about classical releases. If the different > recordings of the same classical piece of word (e.g. an opera) are > properly stored in MB (by the use of release and medium from the NGS), > would it not make sense to put all of them into the same group? For > example, you would have one group for the opera "Les contes > d'Hoffmann" by Offenbach containing all its versions." > > ...shudder. :) And yet, that logic almost holds together, if you > accept the logic for grouping audiobooks by book, bootlegs by concert, > etc, this almost holds together - yet the concept has nothing to do > with grouping distinct releases (and here, and in the next, both have > any reasonable rationale for doing it killed off by the introduction > of Works in NGS...) Please, don't treat bootlegs and operas the same way. From artist's point of view, there is only one piece of work (concert) in case of bootleg and several different pieces of work in case of opera. Moreover, I believe the artist's POV should be the main criterium of releases grouping. _______________________________________________ Musicbrainz-style mailing list Musicbrainz-style@... http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style |
|
|
Re: Release Groups guideline
by Brian Schweitzer
::
Rate this Message:
On Tue, Aug 4, 2009 at 11:31 AM, Pavel Fedyakov <admin@...> wrote:
Yes, I agree. However, that doesn't work outside of albums and singles. it's exactly these cases where the artist isn't involved in the release (bootlegs, soundtracks, VA compilations, etc.) or isn't actually the performer (classical, musicals, opera, soundtracks, audiobooks, etc.) where the "same album" concept breaks. For opera, musicals, classical, etc, from the "artist" perspective, according to whom MusicBrainz uses as artist (the composer), there is only the singular work, not the multiple performances. Now I'm not arguing that opera, musicals, or classical should actually be grouped by work. I'm only suggesting that we're using quite different (and pretty much contradictory) standards for bootlegs (and, it would seem, soundtracks and audiobooks) from the one we're using for classical, opera, and musical theater, in one case overly grouping, while in the other, imho, correctly keeping different things separate. (Though, apart from bootlegs and the "multiple issues of the same cast album of the same musical" case, we don't actually address any of these in the current guideline, so who knows...) Brian _______________________________________________ Musicbrainz-style mailing list Musicbrainz-style@... http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style |
|
|
Re: Release Groups guideline
by Kuno Woudt-2
::
Rate this Message:
On Tue, Aug 04, 2009 at 03:16:16AM -0400, Brian Schweitzer wrote:
> A line by line glance before bed, a few issues, apart from the bootleg > question, though I'm sure there's still issues here I've forgotten :P [snip] > The terminology here is correct for now, but will be incorrect come NGS. > Actually, this entire line becomes redundant under NGS, as the whole > multi-disc problem should likely move to the release and medium pages... Not particularly relevant to the current proceedings, we don't have NGS yet. So this is not an issue which needs to be addressed at this point. I have similarly removed all other NGS related concerns from this reply. > "Multi-disc releases" vs "Box sets" sections - "A special case of the above; > each disc in a box set would normally be part of the same release group. " - > isn't the box set line redundant, esp as it doesn't define anything about > when the discs of a box set would *not* be in the same release group? (or > at least, not until the When Not to Group Releases section) I don't understand your issue with this line, it looks fine to me and not something which needs to be changed before making this official. > "Promotional and pirated versions" vs "Promotional and bootleg versions" - I > know navap doesn't like the word "pirated" being used here; I switched the > wordage from "bootleg" to avoid the "what is intended" confusion problem > mentioned in the bootleg RG edit. ("bootleg" here meaning "pirated copies > of official releases", not "bootleg" in the MusicBrainz standard general > sense of pirated copies, 'real' bootlegs, and homebrews all being "bootleg" > for release status.) In IRC conversation, however, neither of us could > figure a good wording here that would keep "bootleg" in there somehow, > clarify that it only means pirated copies, not ALL bootlegs, and yet not > overcomplexify the wording. I can somewhat understand the dislike for pirated being used here, but I think the intent is quite clear like this, so, again, not an issue which needs to be changed before making it official. Obviously suggestions to rephrase this are welcome. > "Weezer's "Weezer (Red Album)" has ten editions in the database, some > releases from different countries, some deluxe editions, and one > transliteration." - this wording will become out of date the moment someone > adds a second transliteration... Sticking "(currently)" in front of that fixes that I'd say :) > "Mozart's 9 volume, 170 disc "Wolfgang Amadeus Mozart: Complete Works" box > set" example - this RG actually, in of itself, is problematic... It's > grouping 9 different box sets, each individually released before they were > released as a box set. This example (and RG) actually seems pretty much in > direct conflict with the What Not to Group line "A Series of different > volumes that may have been released over time.", given that while yes, the > box was released all at once, the releases in it are titled as they were in > the volume box (and actually in the big box as well), with individual Volume > identification, not identification as a whole box without separate volumes. > Either > http://musicbrainz.org/release-group/314eb0c2-4998-3a8b-a2b7-03ee4e8338b1.html, > or one of the Brilliant Classics "Historic Russian Archives" sets > would > seem to make a more correct and clear example. (Navap, I'd suggest that the > edit to merge all the separate volumes of the box set maybe should be > reverted, if the guideline stays written as it is now...) I agree this is against the current proposed guideline. However, to me it would be more useful to group them like this, so I would be in favour of changing the guideline to allow this particular case. But I guess the rules of when to merge and when not to merge would become rather confusing then. From the POV of getting this guideline official however, it is 'just' an example and can be simply removed. The Enya example is clear enough for the current box-set rules. > "For Box Sets or Multi-Disc Releases where each disc has been released > separately. If these individual discs have not been duplicated due to the > rules outlined at "What Defines A Unique Release", they should not be > grouped together as per the normal box set rules noted above." is a little > awkward to parse; something like: > "For Box Sets or Multi-Disc Releases where each disc has also been released > individually, the individually released releases should not be grouped > together. Instead, the releases should be duplicated for the box set > listing, according to the guidelines outlined at "What Defines A Unique > Release", and grouped according to the normal box set rules." > would be clearer. Agreed. I would further shorten 'individually released releases' in your sentence to 'individual releases'. > "If you are unsure about other cases, please feel free to come on IRC, use > the mailing list, or add a note to the discussion page. " - is this not true > for to any guideline (and thus redundant) - plus, the discussion page or IRC > is not, imho, the place people should be going to for style discussions for > the inevitable cat corner cases... Redundancy in itself is not neccesarily a problem, only if it makes the guideline itself overly long and less clear by restating stuff which has already been said. A guideline should be clear and concise. I see no particular problem with this sentence, other than 'the mailing list' not being accurate (we have several). I also wouldn't object to removing it though, if others agree that this line shouldn't be in it. > "In the case of re-releases with different titles, the principle of > Consistent Original Data" - Given two releases of the same album, with > entirely different titles, I'm not sure how COD is applicable; if our intent > is that the original release's title has greater predominance, we should > just say that. The intent seems to be to prefer the most common title. I'm not sure if that works for all cases, but it should be fine for the majority of release groups. I agree this bit should be rewritten. > "It is OK for a bonus disc EP, remix, live disc etc to be inside the same > "Album" release group as the accompanying album disc. " - See above for how > this breaks under NGS; there is no release "type", so all of these would > have their type wiped away into whatever the RG type is. However, I cannot > think of the correct case, per the RG guidelines, where any of these would > be correct to be grouped with an album, anyhow - so is this line simply > redundant? I think this line is needed to make it clear to those reading the guideline that it is indeed intended to group this discs in the same RG. > * There is no discussion of when to/not to group soundtracks. > > * While there is a mention of 2 different releases of the same musical+cast, > I'd suggest that there should be an example specificly stating that "My Fair > Lady (Original London Cast)" and "My Fair Lady (Original Broadway Cast)" > should *not* be merged. I would suggest both those situations get a bullet point under "When Not to Group Releases together", could you write these? > * Nothing in the guideline about audiobooks/spoken word. The J K Rowling > listing, at the moment, imho is 99% useless in terms of functionality - it > looks more like it's being used as a pseduo-artist-release "author of book" > AR, not a functional grouping of quite distinct releases. (And grouping by > book breaks the moment you have a 2-in-1 audiobook, anyhow...) Looks like these were merged with one RG per book, that doesn't seem right to me. These are not variations of the same 'album', but completely different readings by different people. I would say they are against the current guideline, and we would do well to include this situation in the "When Not to Group Releases together" list. > * Not grouping classical / opera / theater / etc by work, but by recording? > (ie, don't make one giant "Mozart's Requiem K 626" RG for every single > release of the Requiem, no matter by whom...) Yes, same thing as the audiobooks more or less. Ok, lots of stuff here, I'm going to make a seperate post with a short todo to basically summarize all of the above. -- kuno / warp. _______________________________________________ Musicbrainz-style mailing list Musicbrainz-style@... http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style |
|
|
TODO: Release Groups (was: Release Groups guideline)
by Kuno Woudt-2
::
Rate this Message:
Me again :) On Mon, Aug 03, 2009 at 10:41:20PM +0200, Kuno Woudt wrote: > As far as I know, we don't currently have an IdeaChampion for the > ReleaseGroups. Is anyone interested in picking that up? I think the > first task would be to change the wording of the bootleg stuff, as apart > from BrianFreud there mostly seems to be consensus on what to do there. As no one else seems interested, I've made a start and incorporated some suggestions from BrianFreuds mail: http://wiki.musicbrainz.org/User:kuno/ReleaseGroups The things which still need to be taken care of: 1) At the bottom of "When Not to Group Releases together" the line which starts with "If you are unsure about other cases, please ..." should either be removed or slightly change to reflect that we do not have just one mailinglist. Any opinions on this? 2) Under "Use of Attributes" we probably need to rephrase this to not refer to ConsistentOriginalData. Any suggestions? 3) I think we have a reasonable consensus on bootlegs, this needs to be added to the guideline. Other than those two issues, it seems we're not yet clear on how exactly to merge certain kinds of releases, I think in most cases people are being a bit overzealous and we should add these to "When Not to Group Releases together". But please let me know your opinions on each of these, if you can suggest a wording to copy/paste into the guideline, even better! :) 4) soundtracks 5) musicals 6) audiobooks 7) classical -- kuno / warp. _______________________________________________ Musicbrainz-style mailing list Musicbrainz-style@... http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style |
|
|
Re: Release Groups guideline
by Aurélien Mino
::
Rate this Message:
Kuno Woudt wrote:
> >> * Nothing in the guideline about audiobooks/spoken word. The J K Rowling >> listing, at the moment, imho is 99% useless in terms of functionality - it >> looks more like it's being used as a pseduo-artist-release "author of book" >> AR, not a functional grouping of quite distinct releases. It's imho 99% useful in terms of functionality :) As far as I know, the ReleaseArtist (and ReleaseGroupArtist) is the book author. So I don't get your "pseudo-artist-release 'author of book'" AR. And we already have the "has written lyrics" AR. >> (And grouping by >> book breaks the moment you have a 2-in-1 audiobook, anyhow...) >> The same way it breaks when you have a "2 in 1" release. Nothing new here. > Looks like these were merged with one RG per book, that doesn't seem > right to me. These are not variations of the same 'album', but > completely different readings by different people. Since the artist is the book author here, it can be considered variations of the same 'album' ('book') since a writer releases books. In the following sentence from the definition of a release-group, just replace "artist" by "writer" and "album" by "book", I don't think there's nothing shocking: "When an artist tells you "We've released our new album", they're talking about a release group. When his publisher says "This new album gets released next week in Japan and next month in Europe", they're talking about the different releases that belong in the release group that the artist told you about." Since audiobooks are cataloged by book writers, they are the main artist, and you should assume that it can make sense from some POVs. > I would say they are > against the current guideline, and we would do well to include this > situation in the "When Not to Group Releases together" list. > > Since the current guideline is only obvious for album, but not for other release types, I wouldn't say that they are against the guideline. Indeed there is nothing in the guideline about it. A release-group is just a "single logical entity" that groups releases. The album example used to explain release-group concept can be interpreted in multiple ways when applied to different release types, e.g. audiobook. Brian has his own definition of a release-group and he's the more vocal of us, but that doesn't mean that the majority of us agree with him. Other people have different minds, and some have nothing against using release-group as a conceptual entity. Just for your information, nobody (maybe KRSCuan since I'm not sure how to interpret his vote change to Abstain) lastly objected to these edits (No votes were to keep the edits open) and you should not ignore the ones who voted Yes: http://musicbrainz.org/show/edit/?editid=10800343 http://musicbrainz.org/show/edit/?editid=10867950 Discussion happened on http://musicbrainz.org/show/edit/?editid=10678956 - Aurélien _______________________________________________ Musicbrainz-style mailing list Musicbrainz-style@... http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style |
|
|
Re: Release Groups guideline
by chabreyflint
::
Rate this Message:
Even if I can't follow Brian in many of his conclusions, he's got a point here. We should try to have some consistency.
Right know we have basically two different cases regarding "main artist" in MB: The ones where the "author" (composer or writer) is considered the main artist (Classical, Musical, Soundtrack, Audiobooks…) opposed to all other cases where the performer(s) is/are considered the main artist. These two groups should folllow different rules (as they are already). But I think we should try to be coherent within each pf these groups.
The latter case (performer as main artist) is the one we're discussing in the "Guiness-Book" debate about bootlegs (edit #10956897). As it seems a majority is in favour of treating bootlegs the same way we're treating all other releases in this group, based on the initial performance, which to me seems coherent.
For the second group, where the writer/composer is considered the main artist, we started to group different readings of the same audio-book in the same release-group. Logically, and by analogy, we then should also group all recordings of "Don Giovanni" or all performances of "My Fair Lady" in the same release group as well. Do we really want this? For my part I'd consider this highly problematic (and messy).
The best way to proceed (thus keeping the guidelines much simpler, with just very few exceptions) would probably be to stick – for both groups – with the identical philosophy and rules: Same performer, same performance, same concert = same group; different performer(s), different performance = do not merge.
Just my 5 cents. And I would vote no on the "Harry Potter" merges now :-)
On Tue, Aug 4, 2009 at 9:06 PM, Aurélien Mino <a.mino@...> wrote:
_______________________________________________ Musicbrainz-style mailing list Musicbrainz-style@... http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style |
|
|
Re: Release Groups guideline
by Brian Schweitzer
::
Rate this Message:
On Tue, Aug 4, 2009 at 3:06 PM, Aurélien Mino <a.mino@...> wrote:
Kuno perhaps said it better than I can, so I'll just echo his comments here. (I'll snip the later bit more on audiobooks, as it's saying the same as here, and I'd again just echo Kuno's comments).
Personally, I think "J K Rowling wrote the lyrics for audiobook Foo" is misusing the AR; the "had book written by" AR I proposed a year+ back, for musicals (where lyrics and book are often separate concepts) would seem a much better AR to use here, if we had it. <snip> just a "single logical entity" that groups releases. This, of course, is where I disagree. This point of view assumes that RGs have zero semantic meaning, thus any group imaginable becomes allowable, unless specifically barred by the guideline. On the other hand, whether you agree with me or not in applying it to bootlegs, a point of view with some variation on "same release/same audio" as a semantic meaning for RGs blocks a good deal of grouping types, even if they're not specifically barred by the guideline. Brian has his own definition of a release-group and he's the more vocal Yes, but it doesn't quite follow that anyone who disagrees with me on bootleg RGs nesessarily agrees with then merging all audiobooks by book, etc... :P Just for your information, nobody (maybe KRSCuan since I'm not sure how Yes, I read through all those edits; it seemed there were other good reasons given for why not to merge audiobooks, and none of the reasons given for why to do it convince me. I also would disagree with the idea that the search engine should somehow become the fallback solution for over-grouping - this especially, "there is still a "releases" page that lists everything, and there's still a perfectly good search function." That seems to say, "if we overgrouped, and your perfectly legitimate release, with a German title, cannot be found, then you're welcome to give up on the "clean artist RG listing" and look through the entire list of all releases (which kind of negates the point of RGs), or you can always just do a brute force search and see if it comes up". If we're grouping things such that users cannot actually find what they're looking for, even when the release does exist in the database, then what's the point of such a "clean" listing? Is the Japanese-only speaking person who's looking for a listing for a Harry Potter audiobook, read/released in Japanese, supposed to now somehow be able to both know and read the book's title in English? Or, i the audiobook any less legitimately titled in English using the title that sold billions of copies in the US ("Sorcerer's Stone"), rather than the British title ("Philosopher's Stone")? If my audiobook has Sorcerer's Stone on the front, why should I be looking inside a RG that doesn't have the same name as the release in my hand? I'm a huge Issac Asimov fan. I've collected hundred of his books, and have a good many in audiobook form as well. Many, many of his shorter stories have been reissued several times, in different audiobook collections. Will we now, then, have one group which is "all audiobook releases of only book foo", another for "all audiobook releases of only book bar", another for "all audiobook releases of only books foo and bar", one for "all audiobook releases of only book bork", one for "all audiobook releases of only book foo and bork", etc? Is this really any cleaner than actually keeping audiobooks separated by actual release? (The various Hitchhiker's Guide to the Galaxy audio dramas, audiobooks, etc all come to mind here as particularly useful if separated, useless if not.) Brian _______________________________________________ Musicbrainz-style mailing list Musicbrainz-style@... http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style |
|
|
Re: TODO: Release Groups (was: Release Groups guideline)
by MeinDummy
::
Rate this Message:
I should probably have posted my reply to 'RFC: Release Groups guideline' here instead:
http://lists.musicbrainz.org/pipermail/musicbrainz-style/2009-August/008247.html http://www.nabble.com/Re%3A-RFC%3A-Release-Groups-guideline-p24861982.html Sorry, Christian (MD) |
| Free embeddable forum powered by Nabble | Forum Help |
