[NNTP] Extension snapshots 2

[Russ Allbery]

Ken Murchison <ken at oceana.com> writes:

> So, do you want to prohibit this unilaterally, e.g.:

> "Servers MUST NOT allow the use of MODE READER after successful
> authentication."

> Or only prohibit it on mode-switching servers, e.g.:

> "Mode-switching servers MUST NOT allow the use of MODE READER after
> successful authentication."

Can we turn this around and say instead that the client MUST NOT send it
after successful authentication?  Then there's no pressing need to say
what the server has to do.

> I'm open to different/additional text if you have any ideas.  Of course,
> all of this is moot if we decide to get rid of MODE READER.  Where do we
> stand on this?

Well, I've said where I stand on it, and Andrew (as another implementor
who's written news software other than INN) feels (correct me if I'm
wrong) that it's necessary to keep.  If someone comes up with a neat way
of letting INN not need it, that's great by me, but I don't have the time
to work on that and my intuition tells me it would be tricky.

> Can you make a formal decision on whether we want this or not, so we can
> have the base doc and extension docs do the same thing?

Sounds like Clive is on the trail of this already as part of figuring out
what to do with transit vs. reader servers.  I don't think it's necessary
to note transit vs. reader on commands that have their own capability
strings.

-- 
Russ Allbery (rra at stanford.edu)             <http://www.eyrie.org/~eagle/>