NODE 464f5827Re: Sandy and I will run a cypherpunks "moderation" experiment in Jan
Against Moderation <antimod@nym.alias.net>Mon, 6 Jan 1997 16:23:31 -0800 (PST)
-----BEGIN PGP SIGNED MESSAGE-----
> Date: Mon, 6 Jan 1997 18:29:51 -0500 (EST)
> From: "Mark M." <markm@voicenet.com>
>
> The problem with making an undelayed, unedited version of the list
> available is that it would increase the burden on toad.com. I
> suppose someone could run a mail exploder that would receive the raw
> version and mail it out to all subscribers. I don't see the
> increased lag as much of a problem. If an automatic moderation
> program is used, the lag should be insignificant for most posts.
Why do you think that load is a problem? I mean most people are not
going to want to subscribe to multipel versions of the mailing list,
right? So why are three mailing lists with a total of ~3,000 users
any more of a load than 2 mailing lists with a total of ~2,000 users?
Hell, I'll even volunteer to run the cypherpunks-raw list on my own
hardware (thus blowing my nym, if it's really necessary) if I can get
the articles as they are submitted in real time.
> This would not only increase the load on toad.com, but would also
> make it more difficult for people who want to receive the moderated
> version and monitor the moderators decisions. Mail filtering can be
> done simply on many mail programs, but checking a list of moderation
> decisions against the mailing list traffic would be just too
> complicated. NoCeM is a nice idea, but most people on this list
> probably do not have the platform needed to run the software.
1. How would this increase the load? I don't see how multiple mailing
lists cause more load if it doesn't mean more subscribers.
2. Even if it did mean more subscribers, why is load a problem? I
mean, even sendmail can easily handle the current load of cypherpunks,
and there are many packages way faster than sendmail (exim should be
particularly good at this kind of load, for instance. Qmail is also
generally way more efficient than sendmail). And don't tell me these
are hard or a pain to install, because I'm willing to set this up and
run it on my hardware if that's what it takes to get an unedited,
unmoderated cypherpunks list.
-----BEGIN PGP SIGNATURE-----
Version: 2.6.2
iQCVAwUBMtGXboCHQnqYPZ9VAQHzcwQArv9i3yiYlqRxqM1zAN/lhS8z9biL1guM
5YQJRGX8MFdh7IxYBkCvsV6r3qfmpfRKJuF/GqZZ0boYfwIF0BRPT3PGV/qoh1IR
5ltGMAaj/k5fpSIxRBk4NdtWR5RhpvMJSdqo7WDNWBuZtYCozno2G8BXwKPkZ5a2
Phqjs68VswE=
=WWIq
-----END PGP SIGNATURE-----
NODE e367d8d9Re: Sandy and I will run a cypherpunks "moderation" experiment in Jan
"Mark M." <markm@voicenet.com>Mon, 6 Jan 1997 18:40:44 -0800 (PST)
-----BEGIN PGP SIGNED MESSAGE-----
On 7 Jan 1997, Against Moderation wrote:
> Why do you think that load is a problem? I mean most people are not
> going to want to subscribe to multipel versions of the mailing list,
> right? So why are three mailing lists with a total of ~3,000 users
> any more of a load than 2 mailing lists with a total of ~2,000 users?
I guess we will have to see how the "experiment" goes. The original list did
say that one of the non-moderated lists would be discontinued. Anyone who
wants to receive the full list but be able to sort according to moderator's
approval will have a difficult time doing this without subscribing to more
than one list. If one of the two non-moderated lists gets dropped, the only
way to do this would be to subscribe to two lists receiving duplicate messages.
> 2. Even if it did mean more subscribers, why is load a problem? I
> mean, even sendmail can easily handle the current load of cypherpunks,
> and there are many packages way faster than sendmail (exim should be
> particularly good at this kind of load, for instance. Qmail is also
> generally way more efficient than sendmail). And don't tell me these
> are hard or a pain to install, because I'm willing to set this up and
> run it on my hardware if that's what it takes to get an unedited,
> unmoderated cypherpunks list.
Bandwidth is a problem. This list distributes about 50 2K messages to 1500
subscribers per day (this is just an estimation, but I don't think it's too
far from reality). This amounts to 150Mb per day. I'm sure compute cycles
aren't too much of a problem. Anyway, I'm just basing most of this on the
original announcement. I don't know to what extent this will effect toad.
Mark
-----BEGIN PGP SIGNATURE-----
Version: 2.6.3
Charset: noconv
iQEVAwUBMtG4gCzIPc7jvyFpAQEPowgAnnAKh+RxpAEBxcxFS5hMjtaWJXOZYiXa
+KRpmdHSZDxJ0N6JtFr9KCk+w9FOmCkgNGa8xo/D7GNMkHOBxMjoamVRhwWhklwu
roiUbPB7pMuTrnEeh2m+csC58wY7DSQjgIm8XrIKcwkDCWWYJ9fk93j1EBKdjTW9
SSrjzzEMwjjPzoWh6x4nyyGW4A/sgzdhTosPjdjHJd7M//SXDkCJ2UUe3jUF+buS
Hz7iiEzsx+9zmwWxE8zcv5b2jBqIOCGjmRmQjvxfLAQwU6HuD80DZMpDaBVUIGwe
qO5CQsv6sQ9LgVwcPHv3viyeFIWO9NXGNdQ9NOI+Ha0Xq1Hy0TlQ5w==
=pq8e
-----END PGP SIGNATURE-----