GitHub user goldjee edited a comment on the discussion: Proposal: a translation guide, with shared principles and per-language conventions
Well, the most problems were with the terms that are used in Superset and should be consistent throughout all the localised strings. Dashboard, filter, chart, measure, dimension, dataset, column, measure, dimension, annotation, annotation layer, annotation template, database connection, SQL Lab and so on and so forth. The backfill processes one string at a time without any context on already translated strings. So, when someone added annotation templates (I think it was that, but don't remember precisely), Claude invented its own term instead of re-using the one that was assigned for annotation layers. That issue had to be addressed manually. There could also be defined the general rules for i18n. For example, it is usually suggested that if a message in the UI consists of only one sentence ending with a full stop, that full stop should be omitted. Language specific rules could also be defined somewhere alongside. Like, in written Russian letter "ё" is commonly replaced with "е". Some people prefer keeping it. But the thing is that the style is picked once and has to be consistent throughout all the prose. Such rule is a good one to keep documented as well. GitHub link: https://github.com/apache/superset/discussions/43562#discussioncomment-18716255 ---- This is an automatically sent email for [email protected]. To unsubscribe, please send an email to: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
