ietf-nntp DEBUG command (9xx)

[Clive D.W. Feather]

Stan O. Barber said:
>> So why not use 9XX for
>> debugging?
> Sure. Write up an extenstion that defines the 9XX codes be used that way if you
> like.
> 
> It does not belong in the base draft.

Stan,

This is an honest request for clarification. I don't understand why you say
that above, when in another message you wrote:

>> there _is no current practice_ - which makes the whole thing somewhat
>> pointless                                                            
>
> Andrew,
> 
> I don't agree. The whole thing is not pointless. It is pointless to try to use
> current practice as a guide if there is no current practice. To me, that's an 
> opportunity to figure out how things should work and codify that. That's what we
> need to do in this case.
>
> I proposed a DEBUG command. Folks don't like that. There has been some
> alternative text that clarifies what should happen when these codes appear.
> That's a reasonable alternative. However, not handling the problem at all is not
> a reasonable approach.

If it is reasonable to add a DEBUG command to solve this issue, why is it
not reasonable to add 9xx codes to the same end ?


However, the main point of this message is to try to clear the air a bit
here. I have seen the following options (these are *not* mutually
exclusive):

(1) x8x codes are for private use and x9x codes for debugging. This is
what is in 977.

(2) x8x codes are for authentication and x9x codes for private use. The
former roughly matches current practice and the latter is harmless (nobody
has reported seeing x9x codes used in practice). Don't talk about debugging
codes at all.

(3) A new number space (I suggested 19x, but 9xx seems to be more popular
and is probably more sensible) is allocated for debugging codes that are
issued before the true response to a command.

(4) A new DEBUG ACK command is defined to mean "the previous response was a
debugging code; now give me the real response".

(5) A new DEBUG ON|OFF command is defined to activate debugging codes.

I think that covers all the cases.

Can I suggest that we hold a straw vote on those 5 suggestions and any
others that I've missed. Something along the lines of, for each one:
(a) would like to see it included
(b) don't want it, but could live with it
(c) seriously dislike the idea
(d) don't care one way or the other

I'm willing to do the tallying if it would help.

[My own view is 1b, 2a, 3b, 4c, 5c.]

-- 
Clive D.W. Feather  | Work:  <clive at demon.net>   | Tel:  +44 20 8371 1138
Internet Expert     | Home:  <clive at davros.org>  | Fax:  +44 20 8371 1037
Demon Internet      | WWW: http://www.davros.org | DFax: +44 20 8371 4037
Thus plc            |                            | Mobile: +44 7973 377646