> If people agree, I'd like to go through the nmh code replacing defensive
> copying on return with a const return value, and of course go through
> the rest of the code fixing any warnings about constness being casted
> away.
>
> I think defensive copying is the right thing in many situations, but not
> for returned data, since it might be a completely unnecessary overhead,
> and it also makes memory management hard to comprehend since the allocation
> takes place in a different part of the code than the freeing.
Disregard that. The issues in the code are more complicated than I
initially thought, and a lot of what's going on is not defensive copying.
But the memory management *does* need to be made more sensible.
Cheers,
- Joel
_______________________________________________
Nmh-workers mailing list
[email protected]
http://lists.nongnu.org/mailman/listinfo/nmh-workers
