Junk Mails

[John Shott]

Raghav:

I fear that we all get a ton of spam these days and that the spammers 
are winning the war against spam filters and the like.  I can tell you a 
bit about my experiences as well as what we are doing ... and what would 
be hard to do ... related to spam.

For starters, this mailing list (labmembers at snf) is far and away the 
biggest list that we maintain and  is moderated. To my knowledge this 
particular list has not had ANY spam messages slip through since we 
began moderating it about 18 months ago.  Moderation of this list is 
important because it's a list that's in a lot of folk's inbox, is named 
on our web, etc.

We also maintain approximately 200 other mailing lists: about 100 of the 
form equipment_name at snf and another 100 of the form 
equipment_name-pcs at snf that are for technical discussion and 
problem/comment/shutdown reports of each piece of equipment, 
respectively.  I'll look at some of our archives to see if I can come up 
with some statistics, but I haven't  noticed a huge fraction of spam 
messages reaching me that come in via these lists in comparison to the 
overwhelming volume of spam that I receive each day simply addressed to 
me ... and I'm actually subscibed to all 200 of those lists.  Moderation 
of 200 lists would be onerous and would likely be unpopular in the case 
of the problem/comment/shutdown lists because of delays in getting 
reports to the staff that can address them.  Limiting postings to only 
subscribers doesn't work because most of you are subscribed as 
name at snf.stanford.edu but send/receive email from elsewhere.  Limiting 
postings to only Stanford email addresses won't work because of all of 
the non-Stanford labmembers.  For a long time, at least, those mailing 
lists were comparatively spam-free ... both because they are typically 
low-traffic and because the e-mail address never appears on a website 
where it can readily slurped up by spammers.

We don't maintain our own spam filters and rely on the stuff that 
Stanford provides.  I know that's not great, but our computer support 
staff is rather minimal and thus far, at least, that hasn't been our 
highest priority.

I'll try to take a look at some of the e-mail archives and see if I can 
generate some more quantitative number as to how prevelant this problem 
is and will also give some thought to what me might do to improve this 
situation without reducing the utility of these lists.

If labmembers have any suggestions, please send them to me privately and 
I'll consider your ideas ... we frequently hear and adopt good ideas 
from labmembers.

Thanks,

John