You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
OCR add-on should be integrated to the automated translations system. At the moment it cannot be done for the new add-ons hence the blocked label. For the time being new translatable string should have translators commends. It would also help if the add-on could be build with a sconscript and not, as it's being done now, manually, but this is a topic for a separate issue.
After it is integrated *.po files for existing locales should be removed from the repository.
Related to nvaccess#4
Related to nvaccess/nvda#2744
The text was updated successfully, but these errors were encountered:
Update:
Translation of add-ons is not formally blocked anymore from @nvaccess's point of view. Even if many add-on integrations in translation system are still pending (look at PRs in mrconfig repo), Windows Magnifier has been integrated successfully recently.
As explained on the mrconfig GitHub page (https://github.com/nvaccess/mrconfig), if you want to have your add-on translated, you need to ensure your add-on has a stable branch and ask for its translation on the NVDA add-on mailing list.
OCR add-on should be integrated to the automated translations system. At the moment it cannot be done for the new add-ons hence the blocked label. For the time being new translatable string should have translators commends. It would also help if the add-on could be build with a sconscript and not, as it's being done now, manually, but this is a topic for a separate issue.
After it is integrated *.po files for existing locales should be removed from the repository.
Related to nvaccess#4
Related to nvaccess/nvda#2744
The text was updated successfully, but these errors were encountered: