Re: Request for help - cleaning spam from the debian-boot mailing list archive

[Frans Pop <elendil@planet.nl>]

On Saturday 09 May 2009, Christian Perrier wrote:
> Quoting Holger Wansing (linux@wansing-online.de):
> > Hi,
> >
> > Christian Perrier <bubulle@debian.org> wrote:
> > > From my own experience, it's fairly easy to "miss" spams in the
> > > lists of messages, so we really needs a few more people (about 1 or
> > > 2, I think) to go through the archives.
> >
> > I want to help here.
> > I will start at April 2009 and go backwards to the past
> > (will document at http://wiki.debian.org/DebianInstaller/SpamClean)
>
> Unless Frans has another advice, I'd suggest concentrating on months
> that haven't had 5 reviews already.

Note that the "review batch job" (including actual removals, updating
"messages to be reviewed" and review statistics) is only run once a week.

This means that if a month in the archive has already had 2 or 3 checks by 
others, it probably makes sense to wait a week or two [1] as there's a 
good chance that some messages will already be removed by then.

I expect quite a few spam messages to already have a few nominations, so 
having 2 or 3 additional nominations may be enough to start getting them 
reviewed and removed.
We should eventually have 5 scans for every month, but it seems smart to 
delay the last 2 for some time to avoid needless work.

[1] One week for messages with 5+ nominations to be included for review 
and one more week for actual reviews by DDs followed by the actual 
removal.