Re: [CODEC] Base64 decode() handling of invalid encoded data

2008-06-25 Thread Julius Davies
+1 for IOException. I've come around on this. Originally I liked garbage-in, garbage-out idea, but I'm +1 to IOException now. The reason I've come around is that I just cannot imagine any Java developer who would be confused or surprised when presented with an IOException in this situation. I'l

Re: [CODEC] Base64 decode() handling of invalid encoded data

2008-06-25 Thread Jochen Wiedmann
On Sat, Jun 21, 2008 at 2:19 PM, sebb <[EMAIL PROTECTED]> wrote: > How should Base64.decode() handle invalid encoded data? I'd vote for an IOException. As the RFC indicates, such characters are more than likely indicators for errors. If anyone requires such a message, it is easy to implement a Fi