Thomas Mueller wrote:
I use a cron job that fetches mails using IMAP from users INBOX/Spam/
folders.
That is possible, but how do you train ham? One solution might be moving
false positives to another special folder, wait an amount of time and
only then move to a folder where it actually should be. Not userfriendly
at all.
A trigger would be harder: p.e. it would be necessary to define how many
triggers can run at once or users could easily DOS-attack your IMAP server.
Yes, of course limits have to be set, but I don't see a problem here.
IMAP is not used for really large userbases anyway and this is an
optional feature anyway.
Adam Kosmin wrote:
I struggled with dspam for over 2 weeks due to its lack of flexibility
concerning user authentication. I switched to Maia Mailguard and could
not be happier. Everything works exactly like I want it to. You might
consider giving it a look.
I don't use the CGI interface, so no user authentication is required, at
all. The dspam signature is in the headers. They just forward their mail
as an attachment to [EMAIL PROTECTED] or [EMAIL PROTECTED] for training.
Thanks for the replies,
Alex
