Labeling this as a bug as the change (imo) severely violates the principle of least astonishment with seemingly no work-around but to change the field yourself. The change makes harmonious sorting of artist names written in different scripts impossible. While the in-script sorting is also arguable useful to the database, its extremely hard to implement harmoniously in software, while Latin sorting is trivial to implement.

For some background: The software used the normal artist sort for very long, but will now use the sort field for the alias. The issue is that they intentionally not follow the same guidelines. The difference is with foreign scripts. Unlike the artist sort order, the alias one is intentionally for the script used.

https://musicbrainz.org/doc/Style/Aliases#Sort_name

The sort name should be in the appropriate form for the alias locale, since different languages sort names differently (for example, use kana for Japanese sort names). As such, the primary alias sort name in the locale of the main entity name does not have to match the entity sort name for artists, and in fact it often will not (given that the sort name guidelines for artists call for Latin sort names).

*

 

Johann Sebastian Bach has the Russian alias “Иоганн Себастьян Бах”, with the sort name “Бах, Иоганн Себастьян”.

*

Микола Віталійович Лисенко has the Ukrainian alias “Микола Віталійович Лисенко” (matching the artist name), with the sort name “Лисенко, Микола Віталійович” (despite the artist sort name being “Lysenko, Mykola”).

• 日本武道館 (Nippon Budokan) has the Japanese alias “日本武道館”, with the sort name “にっぽんぶどうかん”.

Having a field that would have been completely Latin before suddenly change script entirely, making it not really useful for sorting in a multi-lingual environment is to me a serious design failure and probably not be default and absolutely needs to be able to be turned off. This field is, for me, very useful to sort artists in foreign scripts as same as in Latin. This avoids the issue of Cyrillic, Greek, Arabic, Japanese etc. and other scripts coming after after all Latin artists. I think a lot of other users use it like this, as it's been Latin-only for as long as I can remember.

I don't think changing the guidelines either is an alternative, the schema simply does not support listing a native script and a Latin sort order for aliases. The benefit for better sorting for Latin aliases vs. completely breaking sorting for non-Latin aliases is arguable – it could be useful if you have a library with zero non-Latin aliases – but that isn't something the software should assume for you.