On Tue, Jun 6, 2017 at 12:19 AM, Johannes Zeppenfeld <zeppe...@gmail.com> wrote:
> Yeah, I understand that completely... Is there some way to indicate an > error other than throwing an exception? What I would consider dangerous > here is not that the behavior is undefined, but that debug mode throws an > exception (which is reasonable and lets people (or at least me :) think > they can catch it) but release mode silently does something else instead. I > would expect debug mode to fail noisily (e.g., terminate) for anything that > is undefined in release mode. > KJ exceptions are supposed to be treated as "noisy". As a rule I always prefer throwing an exception over aborting, but I design my code such that the error message is never lost. Technically the opt-mode behavior is not "undefined". You may get a garbage value, or an invalid pointer exception, but the code will never segfault or anything like that. -Kenton -- You received this message because you are subscribed to the Google Groups "Cap'n Proto" group. To unsubscribe from this group and stop receiving emails from it, send an email to capnproto+unsubscr...@googlegroups.com. Visit this group at https://groups.google.com/group/capnproto.
