Comment # 10
on bug 9263
from Evan Huus
(In reply to comment #9)
> You're referring only to not backporting IF we change the API from returning
> a gint to returning a guint, correct? I think both patches I posted are
> suitable for backporting, are they not? At the very least, certainly the
> first patch to packet-tcp.c is suitable for backporting.
The TCP patch is definitely backportable, and the guint idea is definitely not.
Changing the return value in a certain case is iffy, but I think it counts as
an API change in that code *correctly* depending on the old behaviour would
have to change to accommodate the new behaviour. I admit that it would be nice
not to have to worry about these cases anymore even in 1.8/1.10 so I won't
argue the point any further if you disagree :)
You are receiving this mail because:
- You are watching all bug changes.
