[SC-Help] Re: Parser missed URLs in spam..

[Mike Easter]

Duncan Hill wrote:
> This might be the core then.  I can follow the links quite easily
> from my home connection.  Does SC still push those TLDs into the
> lists that make their way to SURBL if it can't resolve them?

I think that the current status is that urls that make it to the
statistics lists where sc-surbl finds them are the ones which are
/reported/.  If they don't get resolved, they don't get presented for
the option of reporting, so they don't go to the list.

I've expressed here before that I think the parsing reporting options
should be expanded so as to 'relieve' the problem of spamvertisers going
'un-statistic/ed'.

I think that the reporter should be presented with the option to say
whether an url was a spamvertiser, as opposed to an innocent bystander,
whether there is any reporting to the provider or not.  I for one don't
like to notify the majority of providers for spamvertisers for my spam.

I think SC's 'level of responsibility' to the spamsource provider to
notify them of a report calling them a source is very different from the
responsibility of SC's reporter to notify the spamvertiser provider.  I
think the notification of a spamvertiser provider should be optional.
The 'duty' of the SC reporter is to recognize if something is a
spamvertiser or an IB -- and the reporting to the provider should be an
option, not a requirement.

Clearly there is some kind of condition in which the parser doesn't want
to or can't provide an avenue for reporting a spamvertiser.  That
shouldn't be a reason or prevention for the statistics process.

An analogy is the situation where there is a spamsource which doesn't
have a 'master' to be notified as a provider for a spamsource, but SC's
algorithm allows a devnull process for notifying so that the source will
count toward the SC blocklist.


-- 
Mike Easter
kibitzer, not SC admin