🚀 go-pugleaf

In article <RoFRRB20w165w@mantis.co.uk>, mathew@mantis.co.uk (mathew) writes:
> randy@psg.com (Randy Bush) writes:
>> mathew@mantis.co.uk (mathew) writes:
>>> It causes UUCP systems using slow modems on noisy lines to waste ...
>>            ^ broken
> 
> No, not broken.  *All* UUCP systems on bad lines will handle large postings
> less well than they handle small ones...
> You could devise some system which automatically split up big articles and
> put them together at the far end, but it wouldn't be UUCP.

Nonsense.  One could (and should) devise such a system, but
claiming that "it wouldn't be UUCP" is meaningless.  Usenet news
isn't UUCP, either; it merely uses UUCP as a transport mechanism.
Do news batching algorithms cease to be UUCP because the files
being Copied Unix-to-Unix aren't single, simple messages?  An
automatic chunking algorithm for large messages would be exactly
analogous to the batching algorithms used for small ones.

>> If one limited the world to the lowest common denominator of broken tools,
>> we would still be using stone wheels.
> 
> If you're saying that everyone using UUCP should just fuck off and disconnect
> from the net, might I ask who the hell you think you are?

I can't speak for Randy, but I'm just saying that people
composing news articles can't be asked to make allowances for
deficient pieces of news software which they might not even know
about.  (I shouldn't state this this way; someone will leap in
and bring up the somewhat-related issue of dropping articles with
nonconforming Date: lines for the umpteenth time.)

					Steve Summit
					scs@adam.mit.edu