|
Neohapsis is currently accepting applications for employment. For more information, please visit our website www.neohapsis.com or email hr@neohapsis.com |
Re: Another SPAM doubt
From: Jorey Bump (list
joreybump.com)
Date: Thu Dec 01 2005 - 16:46:38 CST
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Robert Felber wrote:
> On Thu, Dec 01, 2005 at 04:07:04PM -0500, Jorey Bump wrote:
>
>>now on a low-volume site, running SA in a before-queue content filter using
>>spampd.
>
>
> Out of curiousity, how does dspam handle multirecipient mail?
>
>
>>RCPT TO:<foo
bar.com>
>>RCPT TO:<foo2
bar.com>
>
> < OK
>
>>DATA
>
> < OK
>
>>headers:
>>
>>body
>>.
>
> < 4xx|5xx
>
> What, if foo2 doesn't want to use dspam? In that case, he would lose the mail
> still. Or am I wrong?
I don't know about dspam, I'm using SpamAssassin *globally* in a
before-queue content filter. Yes, it rejects the entire message for all
recipients, but that's the point. My users don't even know their mail is
being filtered (and will complain about getting as many as 5-10 spam
messages in a week!). I use a lot of different spam-fighting techniques,
but *all* of them are apparent to the sender (rejections, no
backscatter), so I will hear about any problems that occur (rare, but it
does happen).
I used to quarantine everything, inspect it, then pass on the ham, but
that was a *true* administrative nightmare. This approach adds
complexity to configuration, and lets me enjoy my vacations a little
more. :)
Don't overlook Chris' point about scalability, which cuts both ways. I'm
not an ISP, and support most of my clients down to their desktops, so
eliminating spam and viruses can save me support calls down the line.
But in a large general purpose population, per-user spam categorization
may be mandatory.
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]