ietf-nntp Timeout in a POST (or IHAVE) command

[Charles Lindsey]

In <ylhewrall6.fsf at windlord.stanford.edu> Russ Allbery <rra at stanford.edu> writes:

>The other two alternatives that occur to me apart from INN's behavior are
>to close the connection without saying anything at all, or to send a 441
>posting failed response (or for IHAVE, probably a deferral) and then close
>the connection.

There is the related problem of the server wanting to stop the transfer of
an article from a client (having seen the first 15MB, and there being no
indication that there is a further 15MB to follow ...). Currently, the
only solution is to drop the connection, which is a bit drastic.

Yes, it is agreed that we aren't going to solve that for this draft, but
it would be useful to know what sort of solution we might propose in the
future. For example, if the solution proposed was for the server to send a
"blind" 5xx response (meaning "please stop spewing"), then clients would
have to expect that. Since "blind" responses are one of the solutions
suggested above for the timeout problem, perhaps we should be asking
whether we are likely to make them a feature in some future extension.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: +44 161 436 6131 Fax: +44 161 436 6133   Web: http://www.cs.man.ac.uk/~chl
Email: chl at clw.cs.man.ac.uk      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5