ietf-nntp Response code warts

[greg andruk]

On Mon, Dec 31, 2001 at 12:42:10PM -0800, Russ Allbery wrote:

> How do you think we should mention this?  Note in GROUP or LISTGROUP that
> the response has a different meaning for the other command, or something
> else?  I don't disagree, but I'm not sure the best way of dealing with
> this (I don't think we can change the response codes).

Right, it's going to probably have to be a permanent blemish.  It does bear
mention, since it violates the draft's own rule that "[e]ach response MUST
start with a three-digit response code that is sufficient to distinguish all
responses" and LISTGROUP is supposed to be an example of the right way to
add an extension :(

A note can be added, um, somewhere (up in the Response Codes section?).

  Common response codes with conflicting behavior are to be avoided in
  newly written extensions.  One historical exception is LISTGROUP, which
  adds a multiline variant to the single line GROUP 211 response.

[What this breaks is the ability to use a "dumb" receiver that packs
responses into an incoming queue without referencing the outgoing commands. 
"Don't implement LISTGROUP" is one answer, but that's cheating.]