Hi Mikael,
here is a fix for PR42769, where a type bound call was resolved statically
into a normal procedure call, but using the local procedure (with the same
name) instead of the original (and correct) one.
It turned out to not be OOP specific, and be a pure module loading problem
in the
Am 06.01.2013 12:45, schrieb Janus Weil:
Since it's not a regression, it does not necessarily need to be
backported. However, the sort of wrong-code behavior seen in comment
#24 is severe enough that it might justify backporting. I leave that
up to you.
FWITW, I think it should be backported
Le 06/01/2013 13:06, Thomas Koenig a écrit :
Am 06.01.2013 12:45, schrieb Janus Weil:
I think your patch is indeed ok for trunk.
Thanks, applied as revision 194949.
4.7?
4.6?
Since it's not a regression, it does not necessarily need to be
backported. However, the sort of wrong-code behavior
Hello,
here is a fix for PR42769, where a type bound call was resolved
statically into a normal procedure call, but using the local procedure
(with the same name) instead of the original (and correct) one.
It turned out to not be OOP specific, and be a pure module loading
problem in the way
