Udo Richter wrote:
just to let you know: the patch you attached here, did not solve the
problem! My VDR crashed again.
Just a few minutes too late...
The patch had another bug that re-introduced the original problem once
again. The new copy constructor did not initialize the aux pointer, and
the assign operator consequently free'd its uninitialized value - and
with some luck, thats a valid pointer.
The attached patch does the missing initialization. Hopefully, thats the
last bug. ;)
That'll teach me not to write code when I'm on my way out to the
vdr mailing list