On Wed, 9 Aug 2000, Philip Hazel wrote:
> On Wed, 9 Aug 2000, Dr Andrew C Aitchison wrote:
> > In that case, anyone freezing viruses in the system filter should take note.
> > Defrosted viruses may slip through filters starting with
> > if not first_delivery then finish endif
> Maybe I'm asleep, but I don't follow your scenario. A message arrives;
> first time through the filter it gets frozen. It will sit there until
> the postmaster unfreezes it. Next delivery, the filter will not be
> active.
You mean the filter rules are not active so the message gets delivered ?
That is what I thought and what worried me.
> Previously, the filter would have been activated again, but
> wouldn't have refrozen because of the "don't freeze if manually thawed"
> rule, so I don't see how this changes things.
I'd forgotten about that rule. It means that the filter was already
potentially letting the virus through at the second attempt.
Last time this was discussed, I thought it had been explained (not sure it
was by Phil) that first_delivery was still set, so the virus would still
be blocked. However my tests didn't convince me that it couldn't get
through, so I wasn't actually using freeze after a first_delivery test.
> [I considered it a bug that first_delivery didn't get unset; the second
> time is clearly not the first delivery attempt.]
That is the way I see it too.
Sorry for any confusion.
Dr. Andrew C. Aitchison Computer Officer, DPMMS, Cambridge
A.C.Aitchison@??? http://www.dpmms.cam.ac.uk/~werdna
