[mb-style] Fwd: [mb-users] Cyrrilic

[DonRedman]

Can we please settle this one?

Looking at the album listing of Tchaikovsky, there does not seem to be a  
single album in cyrillic script. Considering the way MB currently works  
(everywhere Albums are presented with the actual artist name, not with a  
t9/13n or script/language-specific alias), it might be _possible_ to write  
his name in cyrrilic, but it makes no sense whatsoever.

I suggest that an official guideline should be issued, that states  
something like the following:

Use the common English transliteration of ArtistNames for artists that  
have releases in multiple languages.
Use the native script for artists that have releases in their native  
script only.

I do not know how to deal with edge cases, in which the artist has _most_  
releases in his native script, but Scheikowski is definitely of the first  
case.

   DonRedman

------- Forwarded message -------
From: "Marco Sola" <marcosola@...>
To: musicbrainz-users@...
Subject: [mb-users] Cyrrilic
Date: Tue, 20 Dec 2005 14:17:11 +0100

Sorry if I bother you again but, does anyone will do something about this?

http://musicbrainz.org/mod/search/results.html?automod=&mod_type=10&mod_type=40&mod_type=28&mod_type=1&mod_type=2&moderator_type=0&moderator_id=20494&voter_type=0&voter_id=20494&artist_id=122653&artist_type=3&orderby=desc&minid=&maxid=

There are almost a couple of tentative change a week, only on this Artist.

IMHO someone should explain why is in this way and write it big on some  
main
page, adding that Picard workaround instruction for tagging.

I can't do it by myself first because I'm not abe to, second because I'm
still against this conversion, at least since we can handle it as it
deserves. (btw, I can't find anymore the *Official Guideline* where it is
said that this Artist has to be written in cyrillic)

Ciao

MArco / ClutchEr2

.

_______________________________________________
MusicBrainz-users mailing list
MusicBrainz-users@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-users



--
Words that are written in CamelCase refer to WikiPages:
Visit http://wiki.musicbrainz.org/ the best MusicBrainz documentation  
around! :-)
_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

[orion-6]

Don Redman wrote:

> Can we please settle this one?
>
> Looking at the album listing of Tchaikovsky, there does not seem to be
> a  single album in cyrillic script. Considering the way MB currently
> works  (everywhere Albums are presented with the actual artist name, not
> with a  t9/13n or script/language-specific alias), it might be
> _possible_ to write  his name in cyrrilic, but it makes no sense
> whatsoever.
>
> I suggest that an official guideline should be issued, that states  
> something like the following:
>
> Use the common English transliteration of ArtistNames for artists that  
> have releases in multiple languages.
> Use the native script for artists that have releases in their native  
> script only.
>
> I do not know how to deal with edge cases, in which the artist has
> _most_  releases in his native script, but Scheikowski is definitely of
> the first  case.
>
>   DonRedman

That language needs tweaking, since as phrased it would apply to nearly
every Japanese artist and a fair amount of Korean and Chinese ones as well.

_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

[DonRedman]

On Tue, 20 Dec 2005 14:45:56 +0100, Orion wrote:

> Don Redman wrote:
>>  I suggest that an official guideline should be issued, that states  
>> something like the following:
>>  Use the common English transliteration of ArtistNames for artists  
>> that  have releases in multiple languages.
>> Use the native script for artists that have releases in their native  
>> script only.
>>  I do not know how to deal with edge cases, in which the artist has  
>> _most_  releases in his native script, but Scheikowski is definitely of  
>> the first  case.
>>    DonRedman
>
> That language needs tweaking, since as phrased it would apply to nearly  
> every Japanese artist and a fair amount of Korean and Chinese ones as  
> well.

Can you explain this? What is the condition these artists fulfill, andhow  
should they be handled? how should the conditions be rephrased so that the  
result is what you expect?

What about this:

IF
  more than 50% of an artist releases in MB are in their native script
  (i.e. Albumtitle and TrackTitle)
THEN
  store the ArtistName in their native script
ELSE
  store the ArtistName in the common English t13n
FI

   DonRedman

--
Words that are written in CamelCase refer to WikiPages:
Visit http://wiki.musicbrainz.org/ the best MusicBrainz documentation  
around! :-)
_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

[g0llum]

Yes, taschikowski is trying to ruin MB for most of us, I'd like it to be
written as tchaikovsky or tschaikowskij haven't made up my mind yet.. :-)

Seriously: Ruaok mentioned diggin into tagging by alias' and adding support
for transliterations as a intermediate solution (not Schema2) yesterday,
maybe he can give a timeline on when this should be possible - if it's not
in the next few weeks (say under one month, I'd suggest we rename all the
cyrillic artists that have a fair amount of international releases back to
the internation (english) transliteration which is currently used as the
sortname (clutcher2 said so that he corrected those from wikipedia). Maybe
we can change the rules and put the cyrillic name into the sortname field
for time being? (or if choose not to, we can always find the correct native
name if we dig the "edit artist name" moderations)

  g0llum


> I suggest that an official guideline should be issued, that states  
> something like the following:
>
> Use the common English transliteration of ArtistNames for
> artists that  
> have releases in multiple languages.
> Use the native script for artists that have releases in their native  
> script only.

_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

[DonRedman]

On Tue, 20 Dec 2005 15:56:54 +0100, Stefan Kestenholz wrote:

> Yes, taschikowski is trying to ruin MB for most of us, I'd like it to be
> written as tchaikovsky or tschaikowskij haven't made up my mind yet.. :-)
>
> Seriously: Ruaok mentioned diggin into tagging by alias' and adding  
> support
> for transliterations as a intermediate solution (not Schema2) yesterday,
> maybe he can give a timeline on when this should be possible - if it's  
> not
> in the next few weeks (say under one month, I'd suggest we rename all the
> cyrillic artists that have a fair amount of international releases back  
> to
> the internation (english) transliteration which is currently used as the
> sortname (clutcher2 said so that he corrected those from wikipedia).  
> Maybe
> we can change the rules and put the cyrillic name into the sortname field
> for time being? (or if choose not to, we can always find the correct  
> native
> name if we dig the "edit artist name" moderations)

I'd suggest that even if the change can be made quickly, the artist name  
should still be changed to the common (English) t13n. All other scripts  
can then be retrieved via language alieses.

However, I doubt that this will be an easy task. The whole fieature would  
only be useful if the album listing can be filtered by language. Is this  
possible with the ArtistPageRedesign you wrote? In this case I'd suggest:  
fix a vague timieline around march 2006 for both changes, and revert to  
common i13n for now.

Whatever the development decision will be, IMO this case needs an official  
ruling _soon_.

   DonRedman

--
Words that are written in CamelCase refer to WikiPages:
Visit http://wiki.musicbrainz.org/ the best MusicBrainz documentation  
around! :-)
_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

[Jan van Thiel-2]

On 12/20/05, Don Redman <donredman@...> wrote:

> What about this:
>
> IF
>   more than 50% of an artist releases in MB are in their native script
>   (i.e. Albumtitle and TrackTitle)
> THEN
>   store the ArtistName in their native script
> ELSE
>   store the ArtistName in the common English t13n
> FI

Hey, you always stress on the fact that you're not a hacker ;)
Seriously: using percentages in a guideline seems wrong. If someone
adds a couple of releases in another language, all of a sudden the
name has to be changed?

Jan

_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

[orion-6]

Don Redman wrote:

> Can you explain this? What is the condition these artists fulfill,
> andhow  should they be handled? how should the conditions be rephrased
> so that the  result is what you expect?
>
> What about this:
>
> IF
>  more than 50% of an artist releases in MB are in their native script
>  (i.e. Albumtitle and TrackTitle)
> THEN
>  store the ArtistName in their native script
> ELSE
>  store the ArtistName in the common English t13n
> FI
>
>   DonRedman

I mean that it's not uncommon for Asian artists to have 90%+ of their
releases titled in English.  For example, look at the discographies of
two of the more prolific Japanese artists over the last 10 years:
http://musicbrainz.org/showartist.html?artistid=10522 and
http://musicbrainz.org/artist/888994e5-6f53-4f53-bced-b20482aea42a.html 
  Both are well under your percentage for native script releases.  Same
thing goes with newer groups:
http://musicbrainz.org/showartist.html?artistid=252612

Those all have at least a little bit of Japanese, but there are other
artists not in the database that don't have a single song or album
titled in Japanese.  That's what I meant by it wouldn't work

As for how to fix it, I don't think it can be fixed if you try to apply
it to all artists.  I've been thinking about it a lot lately and it
really feels like we need to start adding scope to guidelines, making
them only apply to portions of the DB instead of all.  I'll try to
organize my thoughts about that and make a post later...

_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

[Simon Reinhardt]

Don Redman wrote:

> However, I doubt that this will be an easy task. The whole fieature
> would  only be useful if the album listing can be filtered by
> language. Is this  possible with the ArtistPageRedesign you wrote? In
> this case I'd suggest:  fix a vague timieline around march 2006 for
> both changes, and revert to  common i13n for now.
>
> Whatever the development decision will be, IMO this case needs an
> official  ruling _soon_.

How to shoot oneself in the foot...
Please can we stop putting ourselfs under pressure? Users with high
expectations do that enough anyways.
SG5DisasterRelief is not yet on the live server and even after that
needs to settle down. That was one big step to make a big part of the
community (or perhaps only the part which can shout loudest) happy
again. I really don't think "tagging by alias" can happen that quickly.
And it needs some additional thoughts, extension of the RDF protocol,
Picard coding and a lot of testing.
The Tchaikovsky case is not there since yesterday. Nor is the whole "put
artist names in their native lang&script" thing.
We should not let us push by all the users trying to change
Tchaikovsky's name but think about this calm and with a clear mind. We
all have seen where a Schnellschuss* can lead to.

Simon (Shepard)

---
* For the non-german-speakers: a Schnellschuss is a
rash/overhasty/precipitate action. ;)
_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

[orion-6]

Just a quick follow up,
http://www.tkma.co.jp/tjc/j_pop/hamada/disco.html is a nice example of
an artist that names all her songs in Latin with almost all of them in
English, but still writes her name in Japanese.  Any generalized,
percentage based rule would fail in this case.

Orion wrote:

> Don Redman wrote:
>
>> Can you explain this? What is the condition these artists fulfill,
>> andhow  should they be handled? how should the conditions be rephrased
>> so that the  result is what you expect?
>>
>> What about this:
>>
>> IF
>>  more than 50% of an artist releases in MB are in their native script
>>  (i.e. Albumtitle and TrackTitle)
>> THEN
>>  store the ArtistName in their native script
>> ELSE
>>  store the ArtistName in the common English t13n
>> FI
>>
>>   DonRedman
>
>
> I mean that it's not uncommon for Asian artists to have 90%+ of their
> releases titled in English.  For example, look at the discographies of
> two of the more prolific Japanese artists over the last 10 years:
> http://musicbrainz.org/showartist.html?artistid=10522 and
> http://musicbrainz.org/artist/888994e5-6f53-4f53-bced-b20482aea42a.html 
>  Both are well under your percentage for native script releases.  Same
> thing goes with newer groups:
> http://musicbrainz.org/showartist.html?artistid=252612
>
> Those all have at least a little bit of Japanese, but there are other
> artists not in the database that don't have a single song or album
> titled in Japanese.  That's what I meant by it wouldn't work
>
> As for how to fix it, I don't think it can be fixed if you try to apply
> it to all artists.  I've been thinking about it a lot lately and it
> really feels like we need to start adding scope to guidelines, making
> them only apply to portions of the DB instead of all.  I'll try to
> organize my thoughts about that and make a post later...
>
> _______________________________________________
> Musicbrainz-style mailing list
> Musicbrainz-style@...
> http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
>
>

_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

[Simon Reinhardt]

Orion wrote:

> Just a quick follow up,
> http://www.tkma.co.jp/tjc/j_pop/hamada/disco.html is a nice example of
> an artist that names all her songs in Latin with almost all of them in
> English, but still writes her name in Japanese.  Any generalized,
> percentage based rule would fail in this case.

How come I can see here name written in latin chars on the covers then?
_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

[Robert Kaye]

On Dec 20, 2005, at 5:37 AM, Don Redman wrote:

> Can we please settle this one?

Let's.

Here is how this will play out, consider this an official "ruling" on  
the matter:

At the summit and in particular the days after the summit, Stefan,  
Don and myself spoke about this quite a bit. I weighed two options  
for how to fix this [1] and [2] and had a lot of conversations about  
this topic. Then I flew to London and met up with the last.fm crew.  
One of their hot topics of the day was the issue of transliterations  
and that something needed to be done, quick. I outlined the two  
approaches to them, and the alias approach received a warm welcome  
and they said that they could integrate this approach into their  
system easily. The approach with duplicate artists would wreak havoc  
on their system and would be a lot of work for them.

Given all of these pieces of feedback, approach #1 is the most  
sensical approach to fix this problem. The steps we'll need to take  
to make this happen are:

A. Artists (mostly classical, but not asian -- please help me define  
this properly) in a non-latin script will contain the english-
transliterated version of their name in the artist.name and the  
english-transliterated sort name in the artist.sortname field. Much  
like the way we had it a few weeks ago.
B. Artist aliases will then be cleaned up to 1) indicate the type of  
an alias: typo, AKA, transliteration, etc 2) the language and script  
of the alias. What other alias types should we have?
C. Picard (and perhaps even the web site) get a 'preferred language/
script' option that users can set and if an alias in their language  
exists, Picard will use that alias for tagging.

This has the following benefits:

1. MB Tagger users will stop bothering us, since they get what they  
want.
2. Picard users get more flexibility for tagging their tracks and  
Picard now becomes even more important for tagging your tracks  
correctly.
3. Our customers are not greatly impacted.
4. Its not a hideous amount of code.

This has the following drawbacks:

1. This does not address album transliterations. This will need to  
wait until schema II
2. The database does not fully reflect the true name of the artist.  
This also will be fixed in schema II

So, then I would like to ask the style council to:

1. Define to what artists step A applies to.
2. Update the style guidelines to reflect A.
3. Encourage the community to fix current artists in dispute and make  
people aware of these pending changes.

This will be the next server release after SG5DR -- the exact date is  
still unknown, but early February is a decent guess as to when this  
will happen.

Thoughts??


[1] http://sourceforge.net/tracker/index.php?
func=detail&aid=1059830&group_id=19506&atid=369506
[2] Add a duplicate artist with a transliterated name and add AR  
links between the two artists. Then we would need to tweak the artist  
pages to display albums from both artists.

--

--ruaok      Somewhere in Texas a village is *still* missing its idiot.

Robert Kaye     --     rob@...     --    http://mayhem-chaos.net


_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

Robert Kaye wrote:
> 1. MB Tagger users will stop bothering us, since they get what they  want.

I'm sorry to say this, but no, I'm pretty sure they will not. This is not only
problem with few russian classical composers. Look at these artists for example:
http://musicbrainz.org/mod/search/pre/artist.html?artistid=56320
http://musicbrainz.org/mod/search/pre/artist.html?artistid=1231

And I think changing *all* artists to latin script is one big step backward.

Lukas
_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

[Cristov Russell]

> And I think changing *all* artists to latin script is one big
> step backward.
>
> Lukas

Sometimes you have to take a few steps backwards to have enough steam to
make the big leap forward. ;-)


Cristov (wolfsong)
 
The combination is 1, 2, 3, 4, 5.


_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

[orion-6]

Robert Kaye wrote:

> A. Artists (mostly classical, but not asian -- please help me define  
> this properly) in a non-latin script will contain the english-
> transliterated version of their name in the artist.name and the  
> english-transliterated sort name in the artist.sortname field. Much  
> like the way we had it a few weeks ago.
> B. Artist aliases will then be cleaned up to 1) indicate the type of  an
> alias: typo, AKA, transliteration, etc 2) the language and script  of
> the alias. What other alias types should we have?
> C. Picard (and perhaps even the web site) get a 'preferred language/
> script' option that users can set and if an alias in their language  
> exists, Picard will use that alias for tagging.

To be clear - the cleanup would be for all artists, but only the ones
where it made sense for the largest part of their userbase such as the
Russian composers would have the main artist name set to a latin one?

I'd personally like an alias type and preference to indicate "original"
  script/name.  Rather than having to set English or Japanese or Thai
I'd like each to tag to that language automatically, assuming that's
what their name is in.  Care probably needs to go into wording of that,
to seperate cases like with
http://musicbrainz.org/showartist.html?artistid=244551 - an attribute
might be set on 田中れいな to indicate it's the original script for her
name, but it's not her original name
http://musicbrainz.org/showartist.html?artistid=297112 田中麗奈 that's
linked via the "performs as" relationship, if you get what I mean.

For transliteration and translation aliases, each lanugage+script combo
could probably use a "primary" attribute similar to how last.fm has one
image primary for artists pictures -
http://www.last.fm/music/THE+PINK☆PANDA/+images (possibly a bad example,
I think only automods see the "make primary image" link there next to
the images that aren't.)  As an example, with
http://musicbrainz.org/showaliases.html?artistid=175461 "Twelve Girls
Band" would be English Translation, Latin, Primary while "12 Girls Band"
would just have "English Translation, Latin"  Then "Joshi Juuni Gakubou"
would have "Japanese Translation, Latin Transliteration, Primary while
the other 7 variants of that there would lack the primary attribute.

For artist names as opposed to group names the verbage might need to be
changed.  For http://musicbrainz.org/showartist.html?artistid=22051 
"Park Jung Hyun" might be Korean, Latin transliteration, Primary but
"Lena Park" is more accurately described along the lines of English
naturalization, Latin, Primary

Also, for multi-language artist names and aliases, would we be able to
attach all of the languages to on attribute or run into the problem of
having to decide between an unhelpful "multiple languages" and ignoring
some of the languages to pick one as primary?  I'm mostly thinking of
artists like ANTY the 紅乃壱 with the alias "ANTY the Kunoichi" where
having the attributes English and Japanese, Latin Transliteration,
Primary would make sense.  Same for ones like 中ノ森BAND - "Nakanomori
BAND" is best described as "English and Japanese", not one or the other,
nor as "multiple languages"

> This has the following benefits:
>
> 1. MB Tagger users will stop bothering us, since they get what they  want.
> 2. Picard users get more flexibility for tagging their tracks and  
> Picard now becomes even more important for tagging your tracks  correctly.
> 3. Our customers are not greatly impacted.
> 4. Its not a hideous amount of code.

It's a nice interim solution, I like it.

_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

[orion-6]

Short answer: because it's a funny place.

Covers often have both or just the latin.  The obi (not sure what the US
equivalent is, don't think there really is one) usually has the artist's
name in Japanese.  First good example google pulled up is this Pearl Jam
CD http://www.mercadolibre.com.mx/jm/img?s=MLM&f=9001696_8794.jpg&v=O 
where the cover has them in English like normal and the obi has their
name in Japanese.

If you look on her home page that I linked there her name is written in
Japanese in the titles of all the pages, on the information page, and in
the artists list page.  The few Japanese artists that do go just by
their name in latin get written that way - see
http://avexnet.jp/list/domestic1.html for example where Towa Tei and 4-5
others are written in latin.  Her label has one partially like that,
YOSHI, using just one name.

Simon Reinhardt wrote:

> Orion wrote:
>
>> Just a quick follow up,
>> http://www.tkma.co.jp/tjc/j_pop/hamada/disco.html is a nice example of
>> an artist that names all her songs in Latin with almost all of them in
>> English, but still writes her name in Japanese.  Any generalized,
>> percentage based rule would fail in this case.
>
>
> How come I can see here name written in latin chars on the covers then?
> _______________________________________________
> Musicbrainz-style mailing list
> Musicbrainz-style@...
> http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
>
>

_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

[Robert Kaye]

On Dec 20, 2005, at 10:40 AM, Lukas Lalinsky wrote:

> Robert Kaye wrote:
>
>> 1. MB Tagger users will stop bothering us, since they get what  
>> they  want.
>>
>
> I'm sorry to say this, but no, I'm pretty sure they will not. This  
> is not only problem with few russian classical composers. Look at  
> these artists for example: http://musicbrainz.org/mod/search/pre/ 
> artist.html?artistid=56320
> http://musicbrainz.org/mod/search/pre/artist.html?artistid=1231
>
> And I think changing *all* artists to latin script is one big step  
> backward.

Not *all* artists. As I said, there should be exceptions to this, so  
please help me define the rule for who should be changed and who  
should remain as is.

And while it may seem like a step backwards, please remember that we  
are not discarding any valuable data. I'll make some sort of option  
to add an alias to the artist table that preserves the current  
'correct' information. Maybe we need to add another column to the  
alias table that indicates the 'correct' version of the  
transliteration so we can automatically set these for Schema II.

As for unhappy tagger users, we need to move the more discerning  
tagger users to Picard sooner than later.

--

--ruaok      Somewhere in Texas a village is *still* missing its idiot.

Robert Kaye     --     rob@...     --    http://mayhem-chaos.net


_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

[Robert Kaye]

On Dec 20, 2005, at 11:04 AM, Orion wrote:

> Robert Kaye wrote:
>
>> A. Artists (mostly classical, but not asian -- please help me  
>> define  this properly) in a non-latin script will contain the  
>> english- transliterated version of their name in the artist.name  
>> and the  english-transliterated sort name in the artist.sortname  
>> field. Much  like the way we had it a few weeks ago.
>> B. Artist aliases will then be cleaned up to 1) indicate the type  
>> of  an alias: typo, AKA, transliteration, etc 2) the language and  
>> script  of the alias. What other alias types should we have?
>> C. Picard (and perhaps even the web site) get a 'preferred  
>> language/ script' option that users can set and if an alias in  
>> their language  exists, Picard will use that alias for tagging.
>>
>
> To be clear - the cleanup would be for all artists, but only the  
> ones where it made sense for the largest part of their userbase  
> such as the Russian composers would have the main artist name set  
> to a latin one?
>
> I'd personally like an alias type and preference to indicate  
> "original"  script/name.  Rather than having to set English or  
> Japanese or Thai I'd like each to tag to that language  
> automatically, assuming that's what their name is in.  Care  
> probably needs to go into wording of that, to seperate cases like  
> with http://musicbrainz.org/showartist.html?artistid=244551 - an  
> attribute might be set on 田中れいな to indicate it's the  
> original script for her name, but it's not her original name http://
> musicbrainz.org/showartist.html?artistid=297112 田中麗奈 that's  
> linked via the "performs as" relationship, if you get what I mean.

Yes! That is the proper solution for what I was hinting at in my  
previous message. With this field it will be easy to undo this 'step  
backward' when we roll out the proper solution.

--

--ruaok      Somewhere in Texas a village is *still* missing its idiot.

Robert Kaye     --     rob@...     --    http://mayhem-chaos.net



_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

[Simon Reinhardt]

Cristov Russell wrote:

>> And I think changing *all* artists to latin script is one big
>> step backward.
>>
>> Lukas
>
>Sometimes you have to take a few steps backwards to have enough steam to
>make the big leap forward. ;-)
>  
>
I don't think metaphors help us along here. Remember that with every
change of guidelines and data structures we have to change huge amounts
of data - and you cannot always find every piece of data that has to be
changed at once so we have a transition phase. And then when we have a
proper solution we change it all back? Happy editing...

Simon (Shepard)
_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

[Marco Sola]

Firts of all, thank you all for yourmassive and  positive feedback: now I know you take really care of this issue. Let me try to reassume what I understand and throw my purposal.

I really don't want to waste (and I never wanted to) all useful editing done by moderators on asian Artist, which I found proper, but seems indeed that their work pushed us a littlte too further. I think Don pointed out what could be a rationale for temporary solution about the really harming matter, that are (rusian) *classical* composers.

Like it or not, classical music *needs* to follow separate rules even for basical guidelines and they quite successful do it. So maybe we could build another exception: since we can tell for sure that any classical composer has (a lot) more relesase in latin script than in native one, we can just simply rule down that they have to be be Latin.

It could seem unfair and arbitrary but if you think about it it's not so peregrine as it seems and I know some elder moderators that are affirmative to this. Not to mention the majority of users.

And since I personally changed quite all of them to native script (after that Tchaikovskij voting) and I'm subscribed to all of them, if the rule passes I could easily re-edit them and put native version into alias, waiting for a decent server and client solution to come.

Ciao

MArco / ClutchEr2

_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style

[DonRedman]

On Tue, 20 Dec 2005 20:10:35 +0100, Robert Kaye wrote:

>> Robert Kaye wrote:
>>
>>> 1. MB Tagger users will stop bothering us, since they get what they  
>>> want.
>>>
>>
>> I'm sorry to say this, but no, I'm pretty sure they will not. This is  
>> not only problem with few russian classical composers. Look at these  
>> artists for example: http://musicbrainz.org/mod/search/pre/ 
>> artist.html?artistid=56320
>> http://musicbrainz.org/mod/search/pre/artist.html?artistid=1231
>>
>> And I think changing *all* artists to latin script is one big step  
>> backward.
>
> Not *all* artists. As I said, there should be exceptions to this, so  
> please help me define the rule for who should be changed and who should  
> remain as is.
>
> And while it may seem like a step backwards, please remember that we are  
> not discarding any valuable data.

I would add to that (and here the metaphor might actually help) that the  
step forward we have taken, might have been too soon. IMO the Taikovskiy  
case is similar to the disastrous change to SG5. It is less anarchic, but  
has the same flaw: MusicBrainz forms o huge network of diffenent  
interacting components. The ability of one component of this network to  
deal with i18n, has been used as an argument for a step fowrard.
However, the rest of the network (the display logic of the server, the way  
the clients use the data, the still very english-centric way of  
communication, Picard, the user's way of transforming MB data to file  
structures on their (limited) file systems, etc) *are*not*capable* of  
taking this step.

It is really not a matter of "want" but of "can". Therefore, taking this  
step creates a divide between people, who only use components of the  
network that can deal with the step, and people who use components that  
cannot deal with it.

I think that this breach is harmful. I think that we have gone a step too  
far for a good portion of the network that MB represents. I therefore  
think, we should take that step back, help the other components of the  
network to get along, and then move forward again.

Yes, I know that from the point of view of what the server can do, this is  
a step back from what is possible. And deliberately not using the full  
power fo the system is a think people do not like to do.

But IMO asking what the server can do is too shortsighted. I am not saying  
we should always wait for the weakest element in the chain. What I am  
saying is, we should not move forward before we can provide at least one  
working network that can take the step forward as a whole.

Currently we cannot. Therefore we should step back.

I really like the different proposals that were issued in this discussion.  
I especially like the idea to be strict with classical composers and very  
loose with popular performing artists. I take my very tight proposed  
guideline back, and say that editors should decide based on soft arguments  
like the culture of the artis and their surroundings.

(I know that I will be flamed now. I will not be able to answer because I  
have an exam tomorrow. Have a go :-) )

   DonRedman



--
Words that are written in CamelCase refer to WikiPages:
Visit http://wiki.musicbrainz.org/ the best MusicBrainz documentation  
around! :-)
_______________________________________________
Musicbrainz-style mailing list
Musicbrainz-style@...
http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style