On 8/29/05, Konstantinos Margaritis <[EMAIL PROTECTED]> wrote: > On Κυριακή 28 Αύγουστος 2005 12:52, Scott James Remnant wrote: > > It's an interesting question, certainly; to my mind I don't think > > it's any scarier to dump a scary english message or a scary french > > one. The added advantage to translating them is that the user > > might have the skill to know what's going wrong and fix it, the > > disadvantage is that I have to un-translate them when the error > > reports come in. > > There is a simple solution one to this problem, one which I see is > implemented in professional J2EE platforms, and not only there. > > Use unique codes in error messages alongside with the translatable > text, so instead of having just: > > error: xyz: stat > > instead use the following: > > error: E0943: Could not stat file for xyz > > That way both the developers and the users know what is the error. > > I'm pretty sure that with a well defined schema for errors (eg. EXXXX > for errors, WXXXX for warnings, IXXXX for informational msgs. etc), > it would benefit everyone.
This is a good idea, imho, but I have a comment: A list of error codes which is not handled automatically would be a nightmare and is prone to errors. An automatic mechanism for assigning uniqe error codes in a repeatable manner is necessary. (a hash from the string "<filename>:<function name>:<line>", maybe?) -- Regards, EddyP ============================================= "Imagination is more important than knowledge" A.Einstein
