My 2c on the subject... There's certainly some reason behind all these strings. For instance,if the default naming of the document is not localized in the testtool, the software has no way to check it has succeeded in creating a new instance of the document. And on it goes... Somewhere down this tree the question of course comes up, if it makes sense to do all this specific level/window/GUI - for instance to check the "there's a new Java version available" Window in Catalan (g).
So ... issuing just a warning at such points is a questionable proposition. Think of playing Dungeons and Dragons and you have a lamp and you are not sure if it works and anyhow, which room is it again? I support in any case the idea that for first and topten (or maybe toptwenty, but less than everything) these entries must be localized. It is not a such a big lot... With my rosy glasses on I see the generic product and its localized version as nearly orthogonal - nearly meaning, that I would be VERY surprised in there's first and topten fatal errors in any of the versions (bar those that are just starting up). But, I am happy as a lark when proven wrong: we're motivated by the amount of errors found in the RC. One negative side effect of non-localized strings is that the specific testing branch gets roadblocked and the errors behind this roadblock stay undiscovered. regards smo -- Der GMX SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen! Ideal für Modem und ISDN: http://www.gmx.net/de/go/smartsurfer --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
