Spam Actions

[Julian Field]

At 21:51 11/09/2002, you wrote:
> > I chose "*" rather than "#", but done. And no, I'm not going to bother
> > adding a config option just to set the character. I'm sure you can cope
> > with "*" :-)
> >
>I used # rather than * because some filters implement wild card or regex
>matching and the * has a special meaning there. So far as I know right
>now a # doesn't have a special meaning to any filter that I've run
>across. It's something to consider, and yes I'd find that to be a
>trivial edit...

Good idea, I might change it.

> > I appreciate your feature request, but could you possibly word them a
> > little more gently? Remember I do this for nothing, I'm not some mega-corp
> > you can make demands of. Hope you understand.
> >
>Sorry, I didn't mean to sound harsh or pushy.

You've hopefully seen my personal response by the time you see this
:-)

>As an unsolicited testimonial to your efforts some of the places that
>I've deployed MailScanner were using commercial products for spam
>control. They weren't satisfied with the results (too many false
>positives and too much spam getting through) so I installed a
>MailScanner filter in between the commercial package and the rest of the
>email system. MailScanner identified about 30% of the already filtered
>mail stream as being possible spam with very few false positives. As a
>result the number of complaints from their users about Spam pretty much
>disappeared. Those clients are ecstatic over the results, and even more
>so when cost is compared and are discontinuing the use of the commercial
>solution. And we are talking about moderate sized mail volumes,
>something in the 150-180,000 messages a day inbound from the Internet.
>
>MailScanner is handling that kind of load nicely on a dedicated 2x2Gz
>box with 1Gb of memory. Compared to the other scanners that I've
>evaluated (everything from lower end commercial ones, not Brightmail
>it's just too pricey, through the open source variants) MailScanner is
>way out in front. In terms of code quality, features, and robustness
>nothing else that I've looked at comes close.

Thanks for that. Good to hear I'm displacing the commercial guys :) I
wonder if it's worth setting up a "testimonials" page on the website. All
comments on it would have to be attributed, as otherwise everyone will just
think that I wrote them. What does anyone think? I could ask our webmaster
to write me a bit of PHP to automate it too.

Hopefully I'll have the first speed test result for you tomorrow, based on
processing 20,000 messages on a 2x1GHz P3 box with 512Mb of RAM. It's
running version 3 now. I'll pump the same dataset through version 4 in the
morning (UK time). Version 4 has got a lot of testing to be done on it
before I dare release it to anyone...
--
Julian Field                Teaching Systems Manager
jkf at ecs.soton.ac.uk         Dept. of Electronics & Computer Science
Tel. 023 8059 2817          University of Southampton
                             Southampton SO17 1BJ