Proposed patch: cancel channel

Olaf Titz olaf at
Sun May 14 19:50:22 UTC 2000

> > So there is the question: do we need the responses or would it be better
> > to just skip them? The code change in innd would be trivial of course.
> I guess it largely depends on how important keeping statistics are.  I'd
> actually be interested in having statistics for which cancels got an
> article already on spool and which ones just resulted in an entry in
> history.  The current responses don't give that information, but if
> responses are part of the protocol that gives some place to convey that
> information.

It's also more logical to keep the responses (everything in NNTP and
everything in the control channel except shutdown does give a
response). As long as they don't cross long network links, they are
unlikely to slow things down too much anyway.

Extending the response to contain usable info would require to
propagate that info from ARTcancel() through CCcancel(). Looks easy to
me. (ARTcancel currently returns void, very convenient :-)


More information about the inn-workers mailing list