Hi all,
I’ve been thinking about for a while: do we really need to keep maintaining translations of Orekit error messages in an arbitrary number of langages?
I mean nowadays you can translate pretty much anything on the internet. And modifying the N files when adding a new message is a real pain. I think it’s even deterring contributions.
Cheers,
Romain.
I like the idea of helping users understand their problems easily. We did see on different occasions users having difficulties on the forum due to language being English only (and even once or twice users posting on the forum in their native language). I would therefore prefer to keep this feature, but agree it is low priority.
From a developers perspective, perhaps we could have some sort of script to help maintaining the files, for example checking the number and order of messages in all languages and adding the missing entries with our traditional <MISSING TRANSLATION> feature. The script could even update the English file automatically since it can be generated from the OrekitMessages enumerate. This way, developers would just have to update OrekitMessages (which indeed is the absolute minimum) and let the script do the rest. We would also need to remove the testNoMissingFrenchTranslation junit test.
Even if some forum users don’t master English so well, I think it doesn’t mean they need the error messages in their native langage. I believe they can easily translate a sentence or two on their own.
Anyway, if we want to stick to it, yeah some automation would really help for developers. It’s probably easy to write a script with generative AI actually. Still need to be done though…