[antispam-f] Re: AntiSpam 1.58.2
- From: Harriet Bazley <lists@xxxxxxxxxxxxxxxxxxxxxxxxxxx>
- To: antispam@xxxxxxxxxxxxx
- Date: Wed, 02 Aug 2006 01:56:52 +0100
On 19 Jul 2006 as I do recall,
Frank de Bruijn wrote:
> In article <4e46aa6831antispam@xxxxxxxxxx>, I wrote:
>
> > Anyway, if there really are two consecutive line feeds in the header, my
> > suggestion to fix the bug mentioned in the thread 'Problem with header
> > download' by merely looking at line feeds and ignoring carriage returns,
> > won't handle this.
>
> As it turns out, this was just the same problem as mentioned earlier.
>
> Version 1.58.2 is available from http://www.aconet.org/tools/ .
>
> It ignores carriage returns while reading incoming data. This should
> eliminate the problem of a spurious empty line showing up when a block
> of incoming data ends in the middle of an end of line sequence.
>
There still seems to be a problem of some sort concerning body tests.
I'm seeing a constant stream of e-mails downloaded by default that
ought to be deleted by a body test rule that simply checks for the
presence of a double apostrophe - and which *do* show up as triggering
that rule when subsequently run through the Trial window after download!
I inserted additional Reporter instructions into the user test in
question which strongly suggest - even more suspiciously - that lines at
the end of the header are being tested instead of the start of the body
text. But I honestly can't see any difference between the debug logs
of the ones that *are* correctly trapped and the ones that aren't.
(Logs attached by e-mail; not sure how the mailing list treats
attachments.)
--
Harriet Bazley == Loyaulte me lie ==
Dawn: The time when men of reason go to bed.
-- No attachments (even text) are allowed --
-- Type: application/riscos
-- File: logs,ddc
Other related posts:
