Re: Up to date Gmane.io documentation?

era writes:

> That's a pity. Centralized spam filtering was one of the primary
> attractions of Gmane back in the day.

I've never been bothered/helped much by the spam-filtering, but there
are so many mailing lists, I've probably just been lucky to subscribe to
ones that aren't harassed by spammers much.

> (Still, I rediscovered it now because I wanted to look at
> comp.lang.python via Google Groups, and it was positively *horrible.*
> You should take a look and compare if you still wonder or need to
> demonstrate what the benefits of Gmane are really.

Haha, yes, nntp is really, really nice, and - to be fair - it seems that
Google Groups has been given very little priority.

> There is some vaguely stale documentation at
> https://wiki.python.org/moin/CompLangPython.))

If the "wiki" part is to be believed, that should be possible for anyone
to correct?

> Absolutely, but that's the problem, too; there are tens of thousands of
> groups and all of them start with gmane.* so I was afraid I was missing
> out on something.

👍 - I think those three gmane.{discuss,announce,test} are the main
special ones.

> but now all the documentation I can find says to use the opaque (spam
> spam-use-gmane). Does this still work? What does it do?

A quick grep of the Gnus sources makes me guess this checks for
crossposting to gmane.spam.detected, and that group is still updated by
SpamAssassin, so I guess it should still be useful.

Hm, now that I read that group name, maybe I've got a simple scoring
rule on that...

Yeah, my ~/News/all.score has a rule like this:

  ("xref"
   ("gmane\\.spam\\.detected" -5000 nil r))

which lowers the score of any article that has gmane.spam.detected in
the Xref header.


  Best regards,

    Adam

-- 
 "Halleluja og hurlumhej"                                   Adam Sjøgren
                                                       [email protected]