[NNTP] Fwd: Gen-art review of draft-ietf-nntpext-streaming-05

[elwynd at dial.pipex.com]

OK...

Thank you for the clarification... I realise that to an extent this is all an
academic discussion since NNTP is out there and has been doing a sterling job
for many years.  As I said at the outset I am not an NNTP expert and so I was
not aware that so many other commands could be pipelined already.  I think Russ'
proposal which he just copied in another email would probably suit.  Given that
much of this is about documenting current practice, we shouldn't agonise too
much over this, just as long as people are aware that there is a risk.

I do note that in the case of IMAP authentication and frequently the use of TLS
have been built in from the beginning. 

Regards,
Elwyn

Quoting Ken Murchison <ken at oceana.com>:

> elwynd at dial.pipex.com wrote:
> 
> > Maybe I am missing something here, but the original reason for my raising
> this
> > query is that the streaming commands are different from the others.
> 
> Different how?  Many of the commands defined in the base document can be 
> "pipelined" (e.g. STAT, HEAD, ARTICLE, OVER), in the same fashion as 
> CHECK and TAKETHIS.  The only reason for CHECK and TAKETHIS is that 
> IHAVE is a two step command where it requires the client to wait for an 
> initial response before sending the article, therefore it can not be 
> "pipelined".  CHECK and TAKETHIS simply split the two steps into 
> separate commands.
> 
> So, if you have issues with "pipelined" commands, then you have issues 
> with the base NNTP spec as well as IMAP and POP3.
> 
> -- 
> Kenneth Murchison     Oceana Matrix Ltd.
> Software Engineer     21 Princeton Place
> 716-662-8973 x26      Orchard Park, NY 14127
> --PGP Public Key--    http://www.oceana.com/~ken/ksm.pgp
> 


--