spam.el - ham in spam groups

spam.el - ham in spam groups

[Marcelo Toledo]

If I have ham in spam the solution I know nowdays is setting
gnus-ham-process-destinations and moving it to my reclassify group. Like
this:

(setq gnus-ham-process-destinations
      '(("nnml:spam" "nnml:reclassify")))

normaly I've gone through all spam and marked all ham as read, so when I
leave the group all the ham goes to nnml:reclassify, great, I get into
the reclassify group and respool everything and then it should go to the
correct groups.

I dont know if there is a reason for this but I think all this steps
should't exist. If there is ham in spam it must be repooled again and
not moved to another group that has nothing to do with.

And more, it must be spooled but it can't pass through spam-split again,
because if I marked it as ham there is no need test again.

comments?
-- 
Marcelo Toledo
marcelo@marcelotoledo.org
http://www.marcelotoledo.org
Mobile: 55 71 9141-7181

Re: spam.el - ham in spam groups

[Jonas Steverud]

Marcelo Toledo <marcelo@marcelotoledo.org> writes:

> I dont know if there is a reason for this but I think all this steps
> should't exist. If there is ham in spam it must be repooled again and
> not moved to another group that has nothing to do with.

On my nnfolder:Spam group, I have the Group Parameter
(ham-process-destination respool) which respools the article on group exit.

> And more, it must be spooled but it can't pass through spam-split again,
> because if I marked it as ham there is no need test again.

Why *can't* it be ran through spam-split? Your filter has been told it
is ham (I assume you train it on exit) so it should pass through
spam-split without problem. If it ends up in the spam group again it
is just a good thing to retrain the filter. See the redundant
spam-split pass as a check that the training went all right.

I don't think that it is a way of telling the splitting to "*this*
rule shall not be applied to *this* email" - it might be possible by
adding a user defined function in a : construct to check if it should
be sent to spam-split or not, but I think it is more work then it is
worth. You loose some performance on running spam-split but I think
you have to live with it. OTOH, it takes you longer to read the email
then to respool it through spam-split... ;-)

HTH.

/Jonas
-- 
(        http://hem.bredband.net/steverud/        !     Wei Wu Wei     )
(        Meaning of U2 Lyrics, Roleplaying        !  To Do Without Do  )

Re: spam.el - ham in spam groups

[Ted Zlatanov]

On Wed, 16 Jun 2004, marcelo@marcelotoledo.org wrote:

> If I have ham in spam the solution I know nowdays is setting
> gnus-ham-process-destinations and moving it to my reclassify group. Like
> this:
> 
> (setq gnus-ham-process-destinations
>       '(("nnml:spam" "nnml:reclassify")))
> 
> normaly I've gone through all spam and marked all ham as read, so when I
> leave the group all the ham goes to nnml:reclassify, great, I get into
> the reclassify group and respool everything and then it should go to the
> correct groups.
> 
> I dont know if there is a reason for this but I think all this steps
> should't exist. If there is ham in spam it must be repooled again and
> not moved to another group that has nothing to do with.

You can do that if you wish by setting the ham-process-destination to
'respool.

A lot of people (including me) like to "clone" misclassified ham in a
"ham training" folder, which is much harder to do with the mechanism
you propose.  So respooling as an option is OK, but as the exclusive
mechanism isn't.

Generally the spam.el approach has been to allow everyone to do what
they want with moving spam/ham around.  The new spam.el is even more
liberal in this regard.  It has the "default" behavior which moves
spam out of everywhere and ham out of spam groups; "move-none" which
never moves anything; "move-all" which moves spam and ham regardless
of the group classification.  It is complicated but so are Gnus users
:)

Ted

Re: spam.el - ham in spam groups

[Jonas Steverud]

Marcelo Toledo <marcelo@marcelotoledo.org> writes:

> And more, it must be spooled but it can't pass through spam-split again,
> because if I marked it as ham there is no need test again.

I just found what you is looking for:
      spam-disable-spam-split-during-ham-respool t

-- 
(        http://hem.bredband.net/steverud/        !     Wei Wu Wei     )
(        Meaning of U2 Lyrics, Roleplaying        !  To Do Without Do  )