ietf-nntp poll

[Andrew Gierth]

>>>>> "Maurizio" == Maurizio Codogno <puntomaupunto at tin.it> writes:

 Maurizio> But I still do not see the reason to dictate support of
 Maurizio> streaming.

Existing clients do it, because it works with existing servers.
Therefore new servers need to support it (at least in those situations
where current servers do) in order to maintain compatibility.

 Maurizio> And I don't like the idea that a client sends 
 Maurizio> a stream of data, and an old server silently discards it. 

what old server? are there any servers that fail to handle pipelining
of commands (other than the known exceptions of MODE READER, POST and
IHAVE)?

(please, let's use the term "pipelining" both for consistency with
mail and because "streaming" is already in use to describe the
CHECK/TAKETHIS extension specifically)

 Maurizio> It is easier to add a return code to old servers to make
 Maurizio> them conforming,

Don't be ridiculous - the whole point about old servers is that you
_can't_ expect them all to be changed or updated, so you arrange to
interwork with them wherever possible.

-- 
Andrew.