It would be nice if we had better support for transliterated text, so that people can do things like write their search using Latin letters and find something using Greek letters. This would be particularly useful in situations where you don't currently have the ability to write in a particular script (e.g. if you're using a different computer) or if you're doing a tag lookup with files containing transliterated names in the tags.

For some entities we can already add aliases, but not all entities have aliases, and even for those which do, there's no guarantee that someone will have added an appropriate alias, so I think it would be more effective for the search to be smarter about scripts.

An example would be the Greek word ελληνικά (ellinika). If I search for releases using Greek letters, I get three results ( https://beta.musicbrainz.org/search?query=%CE%B5%CE%BB%CE%BB%CE%B7%CE%BD%CE%B9%CE%BA%CE%AC&type=release&limit=25&method=advanced ) but if I search using Latin letters, I only get one ( https://beta.musicbrainz.org/search?query=ellinika&type=release&limit=25&method=advanced ). Ideally both would return the same results.

There are of course quite a few scripts and each has its own issues, so I wouldn't expect us to be able to handle everything straight away (unless there's something like we found for Python in PICARD-542 that can handle a number of scripts). Greek, Cyrillic and Hangul would be good ones to start with, as they're reasonably common in the database and use spaces between words.