On Wed, 19 Feb 1997, Alan Barrett wrote:
> I think it owuld be sensible to have a configurable limit per-transport
> (in smail3 terminology) on the number of addresses and number of bytes of
> command line line arguments passed to each instance of a program that gets
> executed by a transport.
It is easy to add a batch limit - why didn't I think of that? - and I
will do so. Limiting the length of command line is more problematical.
Well, refusing to run the command if the line is too long is easy
enough, but doing what I think you are implying, which is limiting the
number in a batch according to the length of command line just doesn't
fit with the way Exim works. (After all, I didn't conceive it with UUCP
in mind at all.) I could, however, fairly easily provide a limit on the
number of bytes making up the concatenation of all the addresses in a
batch. That seems a bit of a fudge, but it would do the trick here.
These will be positively the last functional changes for 1.60; I want to
freeze the documentation tomorrow. Then I will do the coding - there are
a few other little things to tidy as well.
--
Philip Hazel University Computing Service,
ph10@??? New Museums Site, Cambridge CB2 3QG,
P.Hazel@??? England. Phone: +44 1223 334714
