According to Martin Sevior <[EMAIL PROTECTED]>:

> I can say that I've experimentally determined that pointers pointing to
> bar.c_str() inside foo , point to garbage outside foo. I had to move some
> UT_String 's to member variables because of this.

That's kinda the problem I fixed in the new hashmap that plagued everybody
for a couple of days.


Hub

Reply via email to