response 115 of 184:

Sep 8 13:56 UTC 2002

Just found this item and read it in a sitting and a half (some
distraction by exploding children).

The danger of privacy violations is minimal.  We have *never* stored
identity information on Grex or on any machine on the cyberspace.org
network.  We do believe in keeping this information secure, and we do
not regard any portion of Grex as secure.  Given how much my social
security number, credit card numbers, checking account numbers, and
drivers license number generally circulate around, I don't think having
one of them in Mark's file adds noticably to my personal risk of
identity theft.

Still, the point that we should be minimalist about this is taken.  I
think records should be deleted a few months after a membership is
turned off.  This requires minimal extra effort, and the data discarded
is plainly of very little interest to Grex.

What I consider much more interesting is the suggestion by scg and
others that we don't need the identity information at all.  We'd just
record the user's name and address taking their word for it.  This would
make becoming a member of Grex substantially easier, and would reduce
the treasurer's workload.  Sounds great.  I'm inclined to agree that as
far as restricting net access goes, we no longer need to collect ID.

My one doubt about this relates to voting.  I would like to be confident
that each member is a different person.  I don't want to make it
excessively easy for people to buy several dozen votes for $18 a peice
just by giving a different imaginary name and address for each one.

Most membership organizations don't have the same problem because they
frequently meet face to face.  Possibly we have to require the ID only
for voting, and we don't have to retain it at all.

response 117 of 184:

Sep 8 19:34 UTC 2002

Arguably, we HAVE had the ID requirement prevent untoward use of Grex.
There was the company down south which wanted the membership to get
outbound Internet services, and balked when the staff asked for the ID
of the designated "responsible party" for our requirements.

Speculation was that the company wanted to use Grex for spamming purposes.
We wound up returning their money.

Then there are the people who drop into party and ask why they can't
telnet out, and we tell them that's for members only and ask them why
they just can't telnet direct to their destination since they're telnetting
in, and we never hear from them again.  They are probably vandals.

So no, the policy is not broken.  It prevents Grex from being black-holed.
Without a membership, it's impossible to hijack a mail relay or any of the
other things which are required to do real bulk spamming.  We can only
speculate how many would-be spammers and other vandals have been prevented
from using Grex for their purposes by our policy, but the only answer I
can think of is "pretty much all of them".

response 119 of 184:

Sep 8 21:10 UTC 2002

Hahah.ahahahah.ahaah.ahaa  
BAN LIBRARIES!

response 121 of 184:

Sep 8 22:27 UTC 2002

He's probably assuming that Tritium would have sufficed.

response 127 of 184:

Sep 9 14:46 UTC 2002

Um, yes, there is.  Please reread #115 and then let us know which part of it
you were referring to in your #116.  :)

response 129 of 184:

Sep 9 16:17 UTC 2002

Re #127:  #126 is a response to Russ' #117.  I forgot, though, that Russ
          never replies live -- he probably hasn't seen Jan's post yet.

response 131 of 184:

Sep 9 19:12 UTC 2002

I do not see anything wrong with keeping this information while an 
individual is a member.  Once the membership expires the information 
should be destroyed, even if the member in question intends to renew and 
misses the deadline.  Sure, it means they'll have to provide it all over 
again, but it shouldn't be the treasurer's job to worry about that.  
That's what happens when a membership or an account is allowed to 
expire: you have to provide you info all over again.

I don't know if such a change in practice really requires and amendment 
change and vote, simply a change in style of record keeping.

Second, I'd like to take the opportunity to thank Mark for all his hard 
work.  He didn't have to take on this challenge, and his efforts should 
be better appreciated by all of you.  Constructive criticism is all well 
and good, but give him credit where it is due, and separate him from the 
policy when you dismantle it piece by piece.

response 133 of 184:

Sep 9 20:49 UTC 2002

One solution to the conflict would be to move towards winding up 
outbound telnet access.   With the replacement of Gopher by the Web,
there's not much publically available stuff out there for people 
to telnet to any more.  I can think of the telnet server 
at Weather Underground (rainmaker.wunderground.com) and maybe the 
UM/MSU/Wayne State catalogs are still telnettable.  
 
As another example of telnet fading away:
If I recall correctly, Michigan State intends to discontinue 
telnet access for the campus e-mail system at the end of this school
year.

The current users of outbound telnet could be grandfathered in, 
or something, if you are worried about inconveniencing anyone.
 
This would be somewhat in accordance with Grexian principles, in 
that it would remove a distinction in system privileges between
members and non-members.

response 135 of 184:

Sep 9 21:58 UTC 2002

Regarding #75; Well, how come ID is only required for outbound network
access, then?  *All* users around here get access to compilers, etc.
And the public access kiosks at Columbia allow one to run Java in a
browser; practically the same thing as giving shell/compiler access.

response 137 of 184:

Sep 9 22:27 UTC 2002

You're telling me that public kiosks on the campus of Columbia allow 
anonymous public access and allow saving of files onto them?  Do they 
have cd/disk drives or e-net ports so walk-up users can download data 
they grab from the web, or upload their keystroke loggers?  

And as for our compiler access, we do a pretty good job of securing our 
own machine, but since we can't rely on the rest of the world to do the 
same, the founders felt that we had the responsibility to implement some 
basic measures to reduce the likelihood of Grex being used as a base for 
launching attacks on the Internet.  Therefore, we're not too worried 
about what you might do with programs you've compiled on our own machine, 
because staff watches, and robocop watches, and our software is developed 
and configured to minimize the possibility of serious mischief, so we 
don't feel the need to ask for ID for compiler access, but if you go 
thrashing about on the Internet from our machine, I think it entirely 
reasonable that we should know just who you are.  Even if you only go 
gently tiptoeing about the Internet from our machine, I still think it 
entirely reasonable that we should know just who you are, because we 
don't know in advance what kind of use you'll make of our machine, and 
it's a hell of a lot more practical to prevent a mess than to clean up 
after it.

The simple fact is we're playing a numbers game.  The vast majority of 
users don't have any intention or desire to cause trouble, the same can 
be said of the members group.  If we offered no threshold, no challenge, 
to those who DO wish to cause trouble, then we'd have a much higher 
percetage of them on our system.  Our measures will not prevent any and 
all possible abusers or abuses, but they do a hell of a lot to reduce the 
incidences we actually have to deal with.  We still get at least 20 or 30 
attempted cracks or attacks on any given day.  If we eliminated our ID 
threshold, nobody would even be able to use our system, we'd be so 
overloaded with assholes with just enough knowledge to be dangerous (as 
opposed to the kind we entertain now).

response 139 of 184:

Sep 10 02:41 UTC 2002

>If we eliminated our ID 
>threshold, nobody would even be able to use our system, we'd be so 
>overloaded with assholes with just enough knowledge to be dangerous (as 
>opposed to the kind we entertain now).

Not proven.  And again, there's a middle ground between allowing completely 
anonymous access and making copies of people's drivers' licenses.  Why do
you feel it necessary to argue the extremes, when the reasonable solution
is probably somewhere in the center -- take names and addresses.