Re: More Active Newsgroups?

From: Moe Trin (ibuprofin_at_painkiller.example.tld)
Date: 07/19/05 

Date: Mon, 18 Jul 2005 20:16:23 -0500

In the Usenet newsgroup comp.os.linux, in article
<2870946.zObM1sbzgH@linux.user.and.happy>, CWO4 Dave Mann wrote:

>I noticed after reading through the misc 75K+ posts that probably half of
>them were the usual flames and rants. I unsubscribed from the group. Too
>much BS.

I don't see that much, but as you are going back that far, you probably
ran into some of the sock puppet attacks. The groups (not just c.o.l.m
but also a.o.l, a.l, and others to a lesser extent) have been targeted
but unknown drones. A killfile works wonders for this. A standard
recommendation has been to auto kill anything cross posted to any
advocacy group. On top of that, you then killfile those that disturb
your wah. Using comp.os.linux.misc and alt.os.linux as two examples
along with this group (figures to 18 July 1500UTC):

2005 c.o.l.misc kills a.o.l kills c.o.l kills
Jan 3064 377 1789 670 162 17
Feb 2270 300 1310 221 279 10
Mar 2090 333 1484 608 841 122
Apr 2179 404 1141 293 439 31
May 3053 1027 1205 437 120 4
Jun 2433 1150 1711 727 163 8
Jul 1223 395 1055 332 63 8

Same rules used for all groups, and no significant changes over the period.

>I agree about the Google idea. That is what I just did with a question I
>had about blocking ICMP ECHO. Got the answer, cut and pasted it.

We reject ICMP Type 8 in. We accept ICMP type 0, 3 (certain codes), 4 and
11, and drop everything else inbound. ICMP type 4 may be filtered internally.
My home setup accepts ICMP type 0, 3 and 11, and drops everything else. Pay
your money, take your pick.

>Tennessee still has at least one working Model 526 at the state election HQ.
>I saw it a couple of years ago when I was doing a voting registration
>project with my students (I was teaching HS at the time). The local
>functionary explained it was used to examine the output from questioned
>cards coming from the one county in Tennessee still using the manual punch
>board system.

When I was in California, they were still using the "Vote-O-Matic" system,
which was manually punched cards. I know that Diebold tried to replace this
with their touch-screen systems, and the state barred them after discovering
some things that weren't exactly on the up-and-up. Here in Arizona, I
understand that there is at least one county still using punched cards, but
by in large they have switched to optically scanned ballots. Nice thing
about that is that when you turn in the ballot, it is immediately scanned
and sanity checked, and bounced if you voted for more than the alloted
number of candidates per seat (normally one, but there might be a "pick up
to five" or similar for positions like members of a school board, etc.).
The machine doesn't indicate what is wrong, but indicates only that there
is _something_ wrong. The poll worker watching the machine will explain what
might be wrong, and will offer to replace the ballot so you can try again.
Normally pretty painless, but the last presidential election was a bit
exasperating - state law says that if you are in line when the poles close,
you will be allowed to vote. At some locations, that last person didn't get
to vote until nearly four hours after official closing time.

        Old guy