Pootle listed such errors. I've just tried Transifex and it did not accept a wrong string. AFAIK Linguist prints a warning and will mark the value (even though it will be accepted).
Translators will be always able to break OpenLP, but I think with the current situation they have to do it on purpose. I've only done some quick checks. So, if somebody wants to investigate the exact behavior and limits of out translation infrastructure... In case someone writes some test script keep be aware: %n and %Ln are wildcats from Qt which are used for plural forms. These wildcards are optional. They have no meaning in Pythons string formatting syntax. -- You received this bug notification because you are a member of OpenLP Core, which is subscribed to OpenLP. https://bugs.launchpad.net/bugs/917421 Title: Missing %s in translation results in an traceback and may make some parts of the program unusable Status in OpenLP - Worship Presentation Software: New Bug description: To prevent the problem we should test all translations when updating these, if both source and destination have the same number of replacement markers such as %s, %d, %f and probably others. Also we should keep in mind, that %.2f and % 2s are correct markers. To manage notifications about this bug go to: https://bugs.launchpad.net/openlp/+bug/917421/+subscriptions _______________________________________________ Mailing list: https://launchpad.net/~openlp-core Post to : [email protected] Unsubscribe : https://launchpad.net/~openlp-core More help : https://help.launchpad.net/ListHelp

