On Sun, 2007-12-23 at 14:16 +0100, Oliver Schäfer wrote: > Thanks Kristis! I didn't expect an answer these days... > > > I'd rather we didn't use an asynchronous change mechanism that may > > fail > > without the Scmbug backend knowing it did. > > I agree with you, but I think we can use it synchronously. It will > not send a real email, I'm just streaming to email_in.pl. Would this > be OK for you?
I see what you mean. Still, email_in.pl ends up calling process_bug.cgi that is expected to double-check the input. We'd be getting into a parsing game trying to figure out what failed. I've long come to terms with the need to reimplement what's needed. I'd rather we reimplemented if their APIs are not reusable, and of course let them know about it. However, their new XML-RPC mechanism aims to be a better API, and one you/we should probably look into. It's an API we should aim to use in the future when talking to Bugzilla. And one we could learn from when trying to ... rewrite the crappy custom glue-daemon protocol to use something more reusable that won't be breaking compatibility when new features are added.
signature.asc
Description: This is a digitally signed message part
_______________________________________________ scmbug-users mailing list [email protected] http://lists.mkgnu.net/cgi-bin/mailman/listinfo/scmbug-users
