Uwe Stöhr wrote: > But why do we have a hard string freeze and why is it about 2 weeks? I mean > if we can fix a bug that changes a string, it is worth it. I don't see why > it harms if 3 strings of 5177 strings are not translated to every language. > The benefit of a fixed bug is greater than a translated string. We should > think about the string freeze policy.
If you do not give the translators confidence that their translations will survive to the release, then you basically depreciate their work. Please consider why string freezes are current practice in all FOSS projects. We actually have a very liberal policy already. Big projects (such as KDE) freeze the strings for every major release, so you cannot even fix a major bug if it invalidates the string freeze in a maintenance release. This is certainly too strict for a small project like LyX, but it shows that your statement that "the benefit of a fixed bug is greater that a translated string" is not widely accepted. IMHO string freezes of at least two weeks (which is short enough) should be respected, and we should only make exceptions for very serious bugs. Jürgen
