In general, for existing languages, using crowdin is going to be the best option.
- When there's a technical term that is shared with the operating system (like 'file', 'folder', 'network', 'proxy' and so on), these should use the same terms as the operating system translation.
- When the translations are updated for new source code, the update process guesses when it runs into translations that no longer fit. This creates
fuzzy
translations. It is better to get rid of bad and broken fuzzy translations first. - A missing translation is better than a wrong/misleading translation.
- Focus on the most exposed strings first. You can use the MultiMC log to determine which translations are used where - it complains about missing and fuzzy translations in
.po
files when they are used.
MultiMC supports hotloading of translations from GetText Portable Object
(.po
) files.
The workflow:
- Get a
.po
file from here the translations repository.- Alternatively, get the
template.pot
and start a new translation based on it. - Alternatively, you can also clone the whole repository using
git
.
- Alternatively, get the
- Put it in the
translations
folder. - Edit it with POEdit or a similar editor.
- See the changes in real time. MultiMC reloads the translations whenever something changes in the folder.
- When done, post the changed files on discord, or on github as a pull request.
If you have any questions feel free to hop on Discord and talk to us.