jef@well.sf.ca.us (Jef Poskanzer) (02/22/91)
Ya know, when I came up with this idea I figured there might be
problems with bozos posting garbage just to prove they could do it.
There was an initial flurry of test messages, and there is currently
another one, but overall the hackers have been amazingly responsible.
It just never occurred to me that the major problem would be with
sysadmins, not hackers.
In retrospect, it's clear that self-moderation version 1.00 is indeed
very fragile with respect to sysadmins. Not just deliberate sabotage
as with Rutgers, but simple misunderstanding. I presented sysadmins
with something they had never seen before: a moderated newsgroup
with no moderator's address - which happened to look very similar
to something they had seen all too often: a bozo screwing up while
trying to create a newsgroup. Most of these people had no idea of
what alt.hackers was about, all they saw was the control message.
No wonder that a few of them went ahead and assumed I was a bozo.
If I were to do it over, I think I would make it look to the sysadmins
like a normal moderated group. Basically, I would lie to the sysadmins
as a whole so that the few idiots and saboteurs wouldn't screw things up
for everyone else.
Unfortunately, it's too late for alt.hackers. I'm going to be sending
out monthly newgroups along with the monthly info messages, and that
should help keep the idiot sysadmins in line, but the saboteurs at
Rutgers have gone into defensive mode and are unlikely to reform.
Oh well, it's kind of fun cancelling the offending articles, even
though cancels don't work very well due to B news's broken propagation
algorithm. Maybe with a dozen people sending them out they'll do
better.
---
Jef
Jef Poskanzer jef@well.sf.ca.us {apple, ucbvax, hplabs}!well!jef
"Our gatewaying has nothing to do with the problem and you damn well
know it." -- postmaster@rutgers.eduokunewck@psuvax1.cs.psu.edu (Phil OKunewick) (02/22/91)
In article <23266@well.sf.ca.us> Jef Poskanzer <jef@well.sf.ca.us> writes: >In retrospect, it's clear that self-moderation version 1.00 is indeed >very fragile with respect to sysadmins... > >If I were to do it over, I think I would make it look to the sysadmins >like a normal moderated group... [ monthly newgroup postings ] I second the nomination for jef to be moderator. And, at the same time, I charge him to moderate the group as he sees fit, allowing people to post directly (pending his approval) and cancelling articles that do not belong, like articles that came through an auto-approving gateway. (I suspect Jef will moderate it the same way he does now. Matter of fact, if he used Jef_Devnull@well.sf.ca.us as the moderator's address, with the mailbox being a soft link to you-know-what, it wouldn't surprise me one bit.) > "Our gatewaying has nothing to do with the problem and you damn well > know it." -- postmaster@rutgers.edu Gatewaying, no. Arrogant postmasters, those big fish in the small pond, though... ObHack: I once fixed a Centronics 781 printer with an off-line problem. After it printed every line, it would deselect and force the user to push "select" to get the next line. So, I ran the "on-line" lamp signal through an inverter, and tied it to the select button. When the lamp would go out, it would trigger the select button, and life would continue as normal. You couldn't deselect the printer, but this was not a problem. The problem later turned out to be a bug in the computer's new operating system revision. In a similar note, I fixed another kind of printer with a paper-motion sensor problem in the same way. It used a photo diode/transistor which shone through the paper holes, watching for the beam to keep breaking. This was a little harder, because tying the signal high or low would indicate a jam to the circuit. The paper-drive stepper motor used phase signals that ran similar to the paper holes, close enough to fool the jam circuit. I connected a wire from one of the phases of the stepper motor logic to the paper jam circuit, and the printer was functional again. (Several days later the replacement part arrived, so I could fix it properly.)