ietf-nntp LISTGROUP

[Russ Allbery]

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

> However, I've been thinking further.

> We changed XOVER to OVER and fixed some problems with it.
> We changed XPAT to HDR and fixed some problems with it.

> We have this one nasty wart of the 211 response.

> Why don't we change LISTGROUP to LISTARTICLES and require the 211 response
> to be the same as that from GROUP. Then:
> * implementations that made the GROUP and LISTGROUP 211 consistent just make
>   the two commands the same;
> * all other implementations handle LISTARTICLES as:
>   - do GROUP with same argument, get response
>   - if not 211, forward response and finish
>   - do LISTGROUP with same argument, get response code
>   - if not 211, forward LISTGROUP response and finish
>   - otherwise forward GROUP response followed by the multiline output
>     of LISTGROUP.

> Thoughts ?

I have two basic reactions:

 * We had to rename XOVER and XHDR (not XPAT; we haven't done anything
   with XPAT) because they started with X.  Since we were renaming them
   anyway, we fiddled with them.  If we could have used existing practice,
   I would have argued for sticking with existing practice, warts and all,
   for the backward compatibility benefits.

 * LISTGROUP is not actually that generally useful enough of a command to
   put a lot of effort into reinventing and force client upgrades to use
   the new command.  My intuition is that not many people use LISTGROUP,
   since one can get the same information from OVER and the client usually
   has to send OVER anyway.

Basically, I think this would be putting way too much effort into an issue
that's not actually significant or important to client authors, and we
should just leave well enough alone.  Even if we change this, there are
still other response code warts in the NNTP protocol; cleaning up all
response code warts would be fun to do, but we can't really do that
because it would mean reinventing the protocol.  And short of cleaning
them all up, halfway efforts aren't really that useful.

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