Re: Windows XP optimization tricks

 



That makes me think of something, but no doubt this has all been royally
thrashed through before...

ISP's get paid, as do major Usenet providers. Both provide Usenet access,
and both can refuse to carry certain groups, and presumable also certain
posts. While moderation might not work, if enough providers had a way to
register spam flags from users, a threshold could be used to determine
automatically what messages to delete and not propagate. This could be
abused, no doubt, but at least it would take a major consensus. Most
spammers will get the required anti-votes more quickly than almost any
other poster.

If this idea was considered and found to be unworkable, maybe it's worth
looking at it again. The big Usenet providers especially, do a lot to
preserve Usenet in a workable form for those who pay, and the number of
paying non-spammers vastly outweighs the number of paying spammers, so the
logical economy of the idea is good, if nothing else.

 

Lostgallifreyan wrote:

   It should be based on IP address. A news server should reject any
posts that exceed a preset limit.  Earthlink limits each account to 75
posts, per 24 hour period, per account.  With broad band service, i get
eight accounts, so i could send a maximum of 600 messages in 24 hours.
How many people have a legitimate reason to exceed that in text
newsgroups?  Another method would be to limit the number of newsgroups
an IP address can post to, in a 24 hour period.  It might take hundreds
of years to hit every newsgroup that way. :)


   They other problem is the anonymous mail to news remailers who don't
give a damn who they piss off.

--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

 



I don't think IP's will do it, at least not alone. Too many spammers know
about proxies. It's still worth trying, but I think like mail spam
blockers, there are certain tricks that can be automated and replace a lot
of human vigilance. If the users' spam flag method is set up, we can vote
off an obvious spam, and it gets deleted from online postings and is
retained and propagated only as a spam example. A few key points of its
content (IP, for example, or name, email, key words, routing info in
header, etc) will be compared against future flagged items, and if there
are enough matches, the automated system can set a blocking rule. Then it
should be semi 'intelligent' and take care of itself. If it gets it wrong,
and people start protesting that legitimate posts are failing to show up
(flagged by a similar method to spam, available at the thread top level),
the people who supervise the filter system should look to see if something
got overzealous, and put online those posts that were cached under the
blocking rule. Such caches might exist for all spam, with deletion after
three days. That way, no legit posts would be lost, unless someone really
fell asleep at the wheel.

The advantage of this idea is that unlike moderation it takes little
intervention, has some protection against loss of legit postings, takes
guidance from current effective spam-blocking practise, and allows almost
all posts to be immediately published and propagated.
 

I've heard that some news providers won't carry posts from others they
don't like. I'm not sure how true that is, but if it is, it could help
there, as the small ones who don't care will be blocked by all the big
ones. Court cases might result, but if the big ones can cite good reasons
they'll win. The small ones would have to police their own turf to survive.

 

Lostgallifreyan wrote:

   How long would the proxies last, with all the greedy bastards trying
to use them at the same time?



<http://www.google.com/search?q=RBL+blacklist&rls=com.microsoft:en-us:IE-SearchBox&ie=UTF-8&oe=UTF-8&sourceid=ie7&rlz=1I7GWYA>

Here is the information about my IP address:
http://openrbl.org/client/#24.110.34.161

Some RBL lists have blacklist everything from Earthhlink. I am
whitelised on two, blacklisted on two, and 32 neutrals.



   Or only take filtered peer feeds from the big guys.  I still say that
mail and news server software has to be upgraded to verify the source of
any message, before accepting it.  The extra overhead would still be
less than all the spam that is processed right now.


--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

 



There are lots of them, and each time one wears out that's just more new
stuff for a system to verify. That's the problem with veryfying sources. I
think it's best to assume innocence at first, no matter what, to be sure
that most people can publish quickly.

I agree the software could be updated to check sources better, but it can't
do anything with virgin data, there's nothing to compare with. That's where
the user spam flag idea comes in. That's how the system best learns what is
spam.