ml> well above most known fidonet software, yes... i'm aware of several
 ml> packages that can handle large messages without problems... one tosser,
 ml> in fact, spools to disk based swap files when it runs out of memory
 ml> that it can process in...

 AL> Not a bad idea, although I'm sure that slows things down a bit. 

speed is not always the best answer...

 ml> and i hate to bring it up, yet again, but... if coders were to actually
 ml> follow the specs, we'd not be having these "problems" with message
 ml> sizes... fidonet specifications specifically allow for large messages
 ml> via use of the phrasing "message body unbounded" ;)

 AL> Yes, the specs have always said that, and almost nobody ever read
 AL> that part! ;-)

oh, they read it but they didn't pay attention... instead they interpreted
and/or assigned their own specs... that is and always has been the downfall
of hobbiest coders :(  sadly, i, too, am one of those... at least until i
don't know how many years ago... since then, the code that i've written has
endevored to follow the actual specs as far as it could... the really
troublesome problem that i've faced, personally over the years, is how so
many terminals and users could be handled by machines with only 64K of RAM
and yet, today, it takes a gig or more of RAM to barely service just one
miniscule user :(

)\/(ark
member, FastEcho Beta Team