Spam inoculation mail support?

Spam inoculation mail support?

[Nuutti Kotivuori]

I am wondering if Gnus already supports the draft for spam inoculation
or training mails:

http://www.ietf.org/internet-drafts/draft-yerazunis-spamfilt-inoculation-03.txt

Such a support would probably be rather trivial to implement. The only
part which requires some carefulness is the md5 checksum calculation
if that authentication is used.

And while on the subject (I haven't had time to dig into spam.el
deeply yet) - is there a simple way to have spam.el invoke some
processor only when mails are moved from a spam group to a ham group
or vice versa - eg. living with the assumption that everything has
already been handled and only errors need to be trained.

Thanks,
-- Naked

Re: Spam inoculation mail support?

[Hubert Chan]

>>>>> "Ted" == Ted Zlatanov <tzz@lifelogs.com> writes:

[...]

Nuutti> And while on the subject (I haven't had time to dig into spam.el
Nuutti> deeply yet) - is there a simple way to have spam.el invoke some
Nuutti> processor only when mails are moved from a spam group to a ham
Nuutti> group or vice versa - eg. living with the assumption that
Nuutti> everything has already been handled and only errors need to be
Nuutti> trained.

Ted> Yes, I do that in my setup which is in the Gnus manual as a spam.el
Ted> example.  My mail gets sorted out on the mail server by
Ted> SpamAssassin, and then misidentified spam goes into a "train"
Ted> folder, while misidentified ham goes into a "trainham" folder.  I
Ted> don't do it personally, but you can set summary exit spam- and
Ted> ham-processors that will run when the spam and ham are moved to
Ted> "train" and "trainham" respectively.

I think what he means is that when in a spam group, if he does "B m" to
move the article to a ham group, the ham processor should be called.
(If that isn't what he meant, it's something that I would like to see,
since it fits my workflow better.)  I think that would be related to one
of the functions I was planning on writing, which would allow you to
call the processors on selected articles.  Then you could remap "B m" to
a function that calls the ham processors and then moves the article.
(Or if gnus-summary-move-article uses hooks, you could do that.)

Unfortunately, real life has kept me busy of late.  I'm hoping that I
can get some coding in next week...

-- 
Hubert Chan <hubert@uhoreg.ca> - http://www.uhoreg.ca/
PGP/GnuPG key: 1024D/124B61FA
Fingerprint: 96C5 012F 5F74 A5F7 1FF7  5291 AF29 C719 124B 61FA
Key available at wwwkeys.pgp.net.   Encrypted e-mail preferred.