I think I might have mistreated spam.el...

I think I might have mistreated spam.el...

[Andreas Fuchs]

Hi,

I just found out why I had believed there was no more spam coming to
me. I use spam.el splitting, and I saw many "filing into
nnml+private:spambox" (that's my spambox) messages, but not one mail
appeared there. So now.

This is the relevant content of my .gnus:
,----
| (setq gnus-spam-process-newsgroups '(("nnml\\+private:.*" (gnus-group-spam-exit-processor-bogofilter
| 							   gnus-group-ham-exit-processor-bogofilter))))
| (setq gnus-spam-process-destinations '(("nnml\\+private:.*" "nnml+private:spambox")))
| (setq spam-junk-mailgroups '("nnml+private:spambox"))
| (setq spam-split-group "nnml+private:spambox")
`---- (sorry for the long lines)

After a few days of having an empty spambox, I looked into my
nnml+private active file, and saw this:

       [...]
     412: newsticker
       1: not-a-mailing-list
       [...]
K    222: nnml+private:spambox
       [...]
       1: spambox

Ugggh. Which of these variables is responsible for this group name? Is
it `spam-split-group'? If so, which value should it use? Just
"spambox"?

After that is resolved, my only problem is how to get the articles from
nnml+private:nnml+private:spambox into nnml+private:spambox. mv should
do it (plus faking the right .active article cound, plus rebuilding
nov), right?

Thanks,
-- 
Andreas Fuchs, <asf@acm.org>, asf@jabber.at, antifuchs

Re: I think I might have mistreated spam.el...

[Ted Zlatanov]

On Mon, 27 Jan 2003, asf@void.at wrote:
> I just found out why I had believed there was no more spam coming to
> me. I use spam.el splitting, and I saw many "filing into
> nnml+private:spambox" (that's my spambox) messages, but not one mail
> appeared there. So now.
> After a few days of having an empty spambox, I looked into my
> nnml+private active file, and saw this:
> 
>        [...]
>      412: newsticker
>        1: not-a-mailing-list
>        [...]
> K    222: nnml+private:spambox
>        [...]
>        1: spambox
> 
> Ugggh. Which of these variables is responsible for this group name?
> Is it `spam-split-group'? If so, which value should it use? Just
> "spambox"?

Splitting functions will return a short group name, and the backend
will convert that to the fully qualified group name.  Thus,
spam-split-group should be just "spambox" or whatever you like, and
that will translate into "nnml+private:spambox" automatically.

This is a Gnus convention (nnmail-split-fancy and nnimap-split-fancy),
not something spam.el controls AFAIK.

> After that is resolved, my only problem is how to get the articles
> from nnml+private:nnml+private:spambox into nnml+private:spambox. mv
> should do it (plus faking the right .active article cound, plus
> rebuilding nov), right?

I would enter the group, select all the articles, and move them.
Should be very quick with nnml.

Ted

Re: I think I might have mistreated spam.el...

[Andreas Fuchs]

On 2003-01-27, Ted Zlatanov <tzz@lifelogs.com> wrote:
> Splitting functions will return a short group name, and the backend
> will convert that to the fully qualified group name.  Thus,
> spam-split-group should be just "spambox" or whatever you like, and
> that will translate into "nnml+private:spambox" automatically.
> 
> This is a Gnus convention (nnmail-split-fancy and nnimap-split-fancy),
> not something spam.el controls AFAIK.

Ah, I understand. Thanks for explaining that.

>> After that is resolved, my only problem is how to get the articles
>> from nnml+private:nnml+private:spambox into nnml+private:spambox. mv
>> should do it (plus faking the right .active article cound, plus
>> rebuilding nov), right?
> 
> I would enter the group, select all the articles, and move them.
> Should be very quick with nnml.

That's the problem: I can't enter the bogus group. Telling gnus to enter
it makes gnus enter the "normal" nnml+private:spambox.

Hm, I'll try mv for now (-:

Thanks,
-- 
Andreas Fuchs, <asf@acm.org>, asf@jabber.at, antifuchs

Re: I think I might have mistreated spam.el...

[Andreas Fuchs]

Today, I <asf@void.at> wrote:
>> I would enter the group, select all the articles, and move them.
>> Should be very quick with nnml.
> 
> That's the problem: I can't enter the bogus group. Telling gnus to
> enter it makes gnus enter the "normal" nnml+private:spambox.
> 
> Hm, I'll try mv for now (-:

Renaming the group in the filesystem to something sensible and in the
"active" file was a better idea. It all works again, thanks for your
help, Ted.

-- 
Andreas Fuchs, <asf@acm.org>, asf@jabber.at, antifuchs