Proposed patch: cancel channel

[Forrest J. Cavalier III]

> To:            Olaf Titz <olaf at bigred.inka.de>
> Subject:       Re: Proposed patch: cancel channel
> From:          Russ Allbery <rra at stanford.edu>
> Organization:  The Eyrie

> Should we make a special exception and try sticking this into INN 2.3 at
> the last minute under the deadline?  It *looks* safe, and it's small, and
> this seems important functionality.  Or should I hold off until after the
> release branch and make this one of the first things to go into 2.4?
> 


Nice patch.  But it should not go into 2.3 if you are going
to release 2.3 this week.

I don't think it will break INN, since the changes are really pretty
isolated into that one mode.  But special exceptions are
a bad habit to start.  Also, the patch does not include
innreport changes needed to parse the log line.  (Minor but
it is there.)

Russ, you really ought to take a look at FTP logs of the
snapshots, if they are available, and see how few
people are actually using -CURRENT versions.  There are only
a handful using my HTTPsync service.  Then
read my

   http://www.mibsoftware.com/bazdev/

concerning open source debugging by testing, if you haven't
already.

If INN had regression testing, last minute changes wouldn't
be so risky.