I'd love it if by default:
a) everything were EXACTLY as before,
b) except the assert stops the current unittest, not the entire thing.
This is what's sensible. Beyond that, feel free to introduce
configurability.
Andrei
Sean Kelly wrote:
On Jul 23, 2010, at 8:24 AM, Andrei Alexandrescu wrote:
Sorry for bringing this up again. I still dream of the good ol' assert. Right
now with the latest druntime, assert(expr) goes completely silent. No message
at all. The program exits with errorcode 1, but you have no idea what happened.
In the last thread someone suggested providing an overridable routine that
prints unhandled exceptions during the unittest run. How about that? I'd left
it silent thus far to avoid cluttering specifically formatted unittest results
with runtime-level output.
_______________________________________________
dmd-internals mailing list
[email protected]
http://lists.puremagic.com/mailman/listinfo/dmd-internals
_______________________________________________
dmd-internals mailing list
[email protected]
http://lists.puremagic.com/mailman/listinfo/dmd-internals
