ietf-nntp Extensions and MODE READER

Russ Allbery rra at stanford.edu
Wed Oct 8 09:58:42 PDT 2003


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

> Our extensions in section 8 introduce five new commands:

>     HDR
>     LIST HEADERS
>     LIST OVERVIEW.FMT
>     LISTGROUP
>     OVER

> How should these interact with MODE READER on implementations where the
> latter actually does something? This isn't an area I'm familiar with.

> Possibilities:

> (1) We say that MODE READER is an abomination and these MUST work whether
> or not it's used.

> (2) We allocate each of them to one of the groups:
> - requires MODE READER first
> - forbidden after MODE READER
> - must work at all times.

> At present, I have (2) with all five requiring MODE READER first. This
> seems a non-brainer for LISTGROUP, by analogy with GROUP, but I'm much
> less sure about the others, particularly the LIST variants. Can someone
> with a relevant implementation please comment?

LISTGROUP is widely implemented already and does require MODE READER.

If HDR and OVER are implemented by copying the existing XHDR and XOVER
implementations and making minor changes, which we were trying to support
as much as possible, they too will require MODE READER.

Given that, there doesn't seem to be any benefit served in supporting the
LIST commands without MODE READER, so I think what you have at present is
the right move.

I think it's preliminary to try to push people to get rid of MODE READER
(in part because the easiest way to do that is to switch entirely to a
different port for transit, something that we've not talked about at all
in this draft and which I don't think we should be talking about at least
right now -- we had some discussions about that a while back).

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



More information about the ietf-nntp mailing list