[NNTP] Revisiting POST as a separate capability

[Peter Robinson]

Clive D.W. Feather <clive at demon.net> wrote:

> Peter Robinson said:
> > I see no good reason to outlaw POST-without-READER in either the
> > protocol or the text.  I think that CAPABILITY options should be avoided
> > where possible and (because?) different capabilities should be
> > independent where possible.
> 
> But different capabilities *aren't* always independent.

As far as NNTP is concerned, POST and READER can be independent if we
allow CAPABILITIES to say so.  Some server implementations might be
structured so that POST without READER never arises, but that's no
reason to make this arbitrary restriction in the protocol.

Russ (and others) suggested some rare but realistic situations in which
it could be useful.  So what's the reason for the restriction?  What
harm is is preventing?

Also, it's quite possible that there will be a future extension that
would benefit from a CAPABILITY argument to POST.  (I'm thinking
specifically of an extra parameter to one of the POST response stages
returning the message-id that will be used, or similar.)

Regards,

Peter