Ben --- I'd very much like a change to this patch. Introducing the
'cmd' member of cmd_parms constitutes a change to the API; I really
think these should be discussed on the list before being committed.
This one, in any case, is unnecessary. The cmd_data of the callback
structure already exists
Unfortunately it is also used in set_rlimit. I'm happy to discuss
it, though. It seemed like a useful and harmless addition to me,
and it is back compatible, unlike many other changes that I'd really
like to make.
It strikes me as creeping featurism. The problem is that there was
Ummm... Brian... isn't this patch a bit *big* for 1.1?
rst