Why are new strings pre-translated based on similar strings?

I’ve been meaning to ask this for some time: why do translations for new strings sometimes get pre-populated with content from other strings (with varying degrees of similarity)?

If a translator doesn’t edit these pre-populated values, doesn’t this create the risk of displaying wrong translations on the website? Or are strings marked as “Needs editing” not pulled?

(NB: I’m not referring to the situation where a source string is edited and the old translation is displayed, as well as the source change - that is most efficient indeed).

1 Like

Exactly this. Basically, what this does is try to keep existing strings when a string gets changed a little bit, so you just need to correct it a little and save rather than translating it from scratch. It is a bit too enthusiastic sometimes about what similar is, so it can suggest strange things at times, but it will never get shown on the site until you move it from needs editing to ready :slight_smile:

We activated it so we can do things like fix caps, add missing periods and fix typos without it entirely losing the translators’ work - even if you do still have to reconfirm.