forum

[Discussion] Preferred metadata choices

posted
Total Posts
8
Topic Starter
Noffy
While not in RC, the general trend in ranking maps is to prefer what is most common, most informative, found most often in official sources, and/or most recognizable/iconic. However there is nothing for actually enforcing this outside of goodwill, and currently people can choose to use the most whichever options freely so long as they are the first ranked map of the song.

There should be preference for what's best to use of options given, but I'm at a bit of a loss for how this should be settled and written down and prioritized.

I'm thinking something like (as a guideline):

Metadata used should be clearly recognizable and readily able to trace back to the original song or source. For romanisation of metadata in other languages, following official sources when they are available is preferred so long as it is easily found and commonly recognized.
Thoughts?
Nevo
Sounds perfect to meeeeee
Hivie
big agree
Serizawa Haruki
It's a good idea to actually write it down as a guideline but the wording seems a bit off to me.

Metadata used should be clearly recognizable and readily able to trace back to the original song or source.
I find this a bit weird because that's not always possible for obscure and hard to find songs and doesn't make it explicitly clear that when there are multiple options, the most commonly used or easily recognizable should be prioritized.

For romanisation of metadata in other languages
Other languages than which one? I think those last three words can be left out while keeping the same meaning.

following official sources when they are available is preferred
The way this is written sounds like a duplicate of the rule "Primary metadata sources must be used as references for metadata."

If the intention was to give priority to official romanisations then the relevant allowances need to be removed or adjusted accordingly:

If an artist has provided an official translation for their name, this may be used in the romanised artist field. Official romanisation may be used for the spelling of an artist's name, but the name order must follow the related rule.
If a Unicode Song title has either an official translation or romanisation provided by the artist, either or may be used in the romanised title field.
Maybe it could be something like
If multiple metadata options are available, priority should be given to the option which is most easily recognisable and traceable back to the original song or source. Official romanisations are preferred so long as they are easily found and commonly recognised.
However, I think it's also important to establish what exactly recognisable and easy to find means. For example, a lot of people probably search songs using the latin alhpabet even if they are originally written in a different writing system because that's easier to read and find, this could also include official translations. But that shouldn't mean that official translations should be prioritized if they are not the most commonly used title.
SilentWuffer
while quite specific I'd like to see people's thoughts on BOF contests as a source. The main argument against it is that it isn't a game, but I believe that they're specific enough to be a source for a song.
Topic Starter
Noffy
Ryan, I think you're on the wrong thread.

Opened PR https://github.com/ppy/osu-wiki/pull/6149 using largely what Haruki suggested for wording since I do agree. It being a guideline does mean there's wiggle room if people struggle to agree on which option is indeed the one most common or easy to find and should be discussed per case as needed.

Quite a few people agreed with this change on discord etc too but didn't post in the thread for some reason >_> please post in the thread next time guys!
Serizawa Haruki
Okay but I think the part about translations is problematic because it could be interpreted as "official translations should be prioritized over romanisations" which I don't think is intended because generally both are equally valid. In fact, I don't know really understand what translations are supposed to take priority over. Unofficial translations are not allowed as metadata options anyway, so there is no choice to be made in this regard. If someone wants to use an official translation they are free to do so according to the allowance, but obviously this should be optional. Using either an official or "manual" (e.g. modified hepburn) romanisation is acceptable to the same degree, therefore no preference is needed. I can agree with official romanisations taking priority over "unofficial" romanisations because they are generally very similar and have the same purpose, so going with the title provided by the artist makes sense in most cases. But this cannot be applied to official translations because that way you'd automatically "force" or give precedence to the translation if both options are equally recognisable.

Maybe the wording of the last sentence could be changed to clarify what exactly is meant, but I don't know the term for romanisations that are not officially provided by the artist.
Morrighan

Noffy wrote:

and/or most recognizable/iconic.
Hard agree (if i understand this correctly), sometimes 'official' metadata gets in the way of actually recognizable metadata - i bring this up too often but, just look at beatmapsets/1084585#osu/2268148 (and most other pokemon music from the games tbh). While the romanization of the official japanese title is indeed 'Kessen! N', searching 'Kessen! N' brings very few results on google (in fact, i see one random youtube video of the song, followed up by a wiki of a game called Kessen, instead of any info on the song). Searching 'Decisive Battle! N' however, instantly brings up many more results that refer back to the original song. Aside from this, most of the boss themes are widely recognized as 'Decisive Battle! [placeholder]'.

So basically, if i've understood this correctly, it would allow metadata that isn't 100% official, but yet far more recognizable than the official translation and so on, and if that is the case, i agree
Please sign in to reply.

New reply