ietf-nntp POST responses

Russ Allbery rra at stanford.edu
Mon Dec 31 04:18:25 PST 2001


Charles Lindsey <chl at clw.cs.man.ac.uk> writes:

> No, I think we can risk it without an extension, as explained above.

I don't.  The gain isn't worth it.  Clients should not be retrying
articles automatically anyway; they should present the textual error
message to the user and let them make a decision.  At that point, having a
separate error code isn't that much of a gain.  Temporary failures on POST
just aren't that common; most servers that I'm familiar with implement
queuing on the server side even if the rest of the news system is down, so
it's very rare that one can connect to the server at all and have it not
be able to accept a post.

That, combined with the fact that I don't think we have any business at
all introducing new reply codes for existing commands without a clear
signal from the client that it can handle it, argues strongly against
working on this in my opinion.

I'd file this one in the "minor things that could have been done better if
we were redesigning the protocol from scratch, but which aren't worth
trying to fix at this point" department.

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



More information about the ietf-nntp mailing list