Of course, like all enhanced security measures, SPF removes convenience; for one, using a .forward file to forward mail from one account to another will fall foul of SPF, as the mail's origin (your other account) and its From: address (the sender) typically won't match (unless all your friends are on your local machine anyway). Secondly, putting an email address other than the one your ISP gives you (or one in a domain belonging to whoever runs your SMTP server) could result in your mail being rejected; which is a problem if you have a mail alias at a k3wl vanity domain but don't use that domain's SMTP servers for sending your mail.
Posted by: kv | http:// | Sat Jan 10 07:24:17 2004
i use it too, es muy bien.
Posted by: atom | http:// | Sat Jan 10 08:04:18 2004
i too have been found my spam problems pretty much solved by spam-bayes.
this is really just another hack that won't work. one thing it won't stop is mail from big-dick@cheap-viagra.com
it will stop a lot of legit mail that's taking an unexpected route.
besides, yahoo's domain authentication thing will be coming out soon, and it seems to be a cryptographically strong way of making sure that "from" address isn't spoofed. they claim it will stop spam: it won't. it'll just stop spammers from using yahoo.com (or any other domain using the service) as a from address.
if anyone wants to take some more offensive anti-spam tactics, check out:
http://business-php.com/opensource/cookiejar/
Posted by: cos | http://polydistortion.net/monkey/ | Sat Jan 10 11:47:44 2004
haven't you noticed that spam now comes with a block of random words to try and fool the bayesian stuff?
(remember "spook" for emacs and such? a bit like that, only in reverse)
Posted by: richard | http://mechanicalcat.net/richard/log | Sat Jan 10 22:48:40 2004
I haven't noticed any of those random-word spams actually working (that is, fooling) spambayes though. Nor have the r.a,n.d,o,m p.u.n,c,t.u,a.t,i.o,n mails fooled it.
From what I understand, the spambayes team spent a *lot* of time fiddling and extending the basic bayesian approach (which is not necessarily something that other bayesian filter implementations have do), and I think I'm benefiting from that :)
Posted by: mark | http://donotuselifts.net/ | Mon Jan 12 06:19:09 2004
Why would spammers *want* to fool advanced spamblockers, anyway? They must know the sort of people who would use them would *never* fall for their crap.
Posted by: acb | http://dev.null.org | Mon Jan 12 06:27:39 2004
Either that or are customers of an ISP that uses an advanced spam blocker, but who themselves don't know about it. If you're a spammer, it's not too great a leap to convince yourself that you're a freedom fighter trying to outwit the censors to get the truth out to the public. Just like in Cuba or China or somewhere, only in this case, the "truth" is not forbidden political news but dodgy Viagra pills and financial scams.
Posted by: mark | http://donotuselifts.net/ | Mon Jan 12 15:17:28 2004
Kinda like the trolls on USENET who'll spend all their time dodging killfiles, because THE TRUTH MUST OUT, DAMNIT!
Posted by: acb | http://dev.null.org | Tue Jan 13 03:15:51 2004
Yes, though at least some of the trolls have personalities. The spammers are just scumbags after a quick buck. (Well, except perhaps for the time-travel guy, but he's probably nuts.)
Want to say something? Do so here.
Note to spammers: This comment system applies the rel=nofollow attribute to the poster's URL and all links. Posting links to this page will not improve their search engine rankings.
Please keep comments on topic and to the point. Inappropriate comments may be deleted.
Note that markup is stripped from comments; URLs will be automatically converted into links.
Posted by: richard | http://mechanicalcat.net/richard/log | Fri Jan 9 22:23:01 2004
I've been using spambayes for a few months now. I get maybe a message a day that it isn't sure about, but otherwise I just don't see spam. I've not had a single false-positive.