[Date Prev][Date Next] [Chronological] [Thread] [Top]

Problems in HEAD


HEAD code looks broken; if I go back about 40 hours it works correctly.
There's some dangling reference (fails unpredictably on malloc/free).
I'm trying to trace the problem; Kurt's reworking of the DN stuff should
be OK, while I'm a bit worried about the DN ber scanning as 'o' instead 
of 'a' (and I'm not sure I know what I'm talking about ;-).

I'll try to apply all the changes step by step.