Dave K: Why Spams?

Dave Kristofferson kristoff at NET.BIO.NET
Thu Dec 14 02:18:26 EST 1995


I appreciate Don Chen's words of encouragement.  However, I want to be
clear to everyone that my issue with John Kuszewski's original post is
merely about the tone, not the concerns that he raises.  Spamming has
been an issue that I have had on my plate almost daily for several
months and dealing with it actually delayed the release of our
hypermail archive by about three weeks as we diverted efforts to
enhance the scripts that we use to process messages to our moderated
newsgroups (one recent attempt to combat this problem).

Let me briefly review what we have been doing, but first this
preamble.

I want to point out again that we have about 0.7 FTE's to work on this
project.  Although we do have funding for the project, we are running
close to 100 newsgroups and mailing lists with a very small
committment of staff.  Dave Mack, the BIOSCI systems administrator and
programmer, dedicates about 20% time to the project, Julie Lawrence
and I share the tech support, QA, documentation, and I handle project
admin plus fund raising in the other 50% allocation.  I also donate
time to the project after hours like I am doing right now after 10 PM.
All three of us have other jobs which are our first priorities; so
everyone, please be aware that this project, until it raises
additional revenues, continues to operate very successfully *within
its limitations*.  Anyone who thinks this is a massive bureaucratic
operation with dollars to squander and is unresponsive to its readers
should realize that this operation is actually very lean.  Any delays
that people might encounter in responses to their requests are simply
a result of our being stretched thinly, not due to any lack of desire
to respond.

Enough excuses; what have we been doing about spams?

The primary people who are bothered by spams are our e-mail users.
This was one of our motivations in developing our new hypermail
archive at http://www.bio.net.  People who did not have a local USENET
feed in the past and were stuck with e-mail now have a way to
participate in the groups through their Web browser which requires a
lot less in terms of local support than running a local USENET system.
We continue to hope that this will shrink the mailing lists and thus
reduce the number of spams that end up in personal mailboxes.  People
can scan the message headers in hypermail and just ignore those with
spam-like characteristics.

We also recently wrote the BIOSCI miniFAQ which addresses the issue of
spams in its very first question.  Of course, our answer there will
not make John happy since we basically told people that protesting is
usually useless, and we think that people should just ignore spams.
For example, hardly a day goes by when my friendly postman does not
stop by my house with the latest supply of junk mail.  Do I pull out an
Uzi and blow him away so I don't get any more junk mail?!?  Of course
not.  I just file the stuff in the trash can.  Unfortunately, the
growth of commercial entities on the net means that the days of purely
research e-mail is over and this problem will be with us for the
foreseeable future.  Unlike many LISTSERVs that allow anyone to
retrieve their mailing lists, we have always kept the mailing lists
here protected precisely because we did not want people taking lists
for commercial purposes.  The wisdom of this policy is finally
becoming clear with a vengance.

Many people have suggested writing filters.  This is possible and some
are available.  My feeling to date (and this can be debated) is that
(1) filters only work on preselected text and thus are easy to get
around (true we could nuke everything with Subject "MAKE MONEY FAST"
but it take all of two seconds to change the Subject: to "Major
breakthrough in xxx research" and leave the body of the message the
same as before, (2) filters often nuke messages that should be
distributed (should we kill every message that mentions the word
"money"??), (3) our resources are limited, so (4) is this really the
best way to expend them? So far my answer has been "no."

Instead we have encouraged groups that are bothered by spams to
consider moderation.  Although the moderation mechanism can be easily
overridden on the newsgroups, we are decoupling the mailing lists from
the moderated newsgroups and have written software which allows only
the newsgroup moderator to post to both the newsgroup and the mailing
lists.  This software was completed in November and is in use by 10-15
of our moderated groups.  All the moderator has to do in a nutshell is
read mail and forward it on to an approval address here for
distribution; very little labor is involved beyond the reading of the
message.  Once again, we recognize that we have less control over the
USENET distribution, but anyone with a newsreader can simply not
select to read messages that have Subject: MAKE MONEY FAST, for
example, or make a kill file.  Our main focus was to protect our
e-mail users.

In summary, in the last two months we have put up a new means for our
e-mail users to avoid spams by getting off the mailing lists and using
hypermail, and by protecting the lists on our moderated newsgroups
with new software.

We will investigate John's allegation that our groups have more spam
than others (I still tend to doubt that, but, as I said earlier we may
just retain the residues longer if cancelbots launched elsewhere are
not also cleaning up our groups; Dave Mack and Cornelius Krause so far
have both indicated to me that cancelbots are run net-wide as was my
understanding).  We may implement a means to run cancelbots on the
bionet groups, but also have to confront the issue of who will take
the time to select messages for processing by this means.  The BIOSCI
staff can not possibly do this on all of our newsgroups; also
cancelbots, being after the fact "protection", will not protect the
mailing lists on unmoderated newsgroups, so the question remains as to
whether an expenditure of effort on this is really worth it compared
to others items on our list.

We will also continue to consider any good suggestions for resolving
this problem and will act on them as soon as our resources permit.
John may get pissed off because some of his time is wasted by these
items.  I have been working to develop this system for a decade and
we handle tech support complaints about this all of the time.  I have a
far greater incentive to solve this problem than John does, but I also
have to weigh a lot of factors before deciding to commit resources.

Having said all of the above, I am not also going to take additional
time to go into some of the potential legal issues that using
cancelbots might entail!!

John, if you've read this far, I just want to say once again, that it
was not my intention to discourage you from raising the issue once
again.  Please just change the tone of the message in the future and
don't jump to the conclusion that we aren't doing anything.  Dave Mack
and I have thought about this subject at length; you are free to
disagree with our conclusions, of course, but I hope that you at least
understand some of our thoughts now.

				Sincerely,

				Dave Kristofferson
				BIOSCI/bionet Manager

				biosci-help at net.bio.net




More information about the Bioforum mailing list