outgoing internet access for non-members

Mar 14 08:34 UTC 1999

I don't really remember exactly where the discussion of allowing people
who aren't paying members outgoing internet access from grex, but it
seems to have died out without there being any good reason that I remember.

Anyway, I'd like to make this motion:

   Outgoing internet access from grex will be allowed to any Michigan
   resident who provides suitable proof of identity.  The exact definition
   of `suitable proof of identity' will be determined at the discression of
   the board of directors.

response 2 of 127:

Mar 14 18:58 UTC 1999

re: 0
        Are you sure that's a good idea?

        You've got 70+ folks incoming on telnet, ftp and lynx. Talk,
        party and whatnot..

        Are they supposed to be "outgoing" with Telnet? Lynx? IRC?

response 4 of 127:

Mar 15 00:32 UTC 1999

As one of the people who would benefit from this, I suggest that we _not_ do
it.  Grex does have ongoing financial needs.  And far fewer people abuse any
service when they have to pay a nominal fee for it.  I see no reason to
suddenly change Grex into a free service by unlinking the two most
"abuse-prone" commands from membership. 

The benefit is so minor in comparison with the new demands on volunteer time
that would ensue, that I think we should leave them as that small nudge that
helps people decide to support Grex.  

response 6 of 127:

Mar 15 03:18 UTC 1999

I'd like to see the only difference between members and non-members
be the right to a membership vote.  I really don't like the idea
of tiered access, where those who pay get perks for their money.
If the system resources simply can't handle our being able to
offer something to everyone then I'd rather it not be offered at
all.  But I know this is not the way most folks see it.  I also
suspect that if Grex were designed today, by the existing members,
that it would probably look a whole lot different than it does.
It's real tempting to go the safe route, looking for the worst
in folks and being unwilling to take chances, and looking toward
membership rewards as a means of keeping our finances comfortable
and our club secure.

I think we should try opening outgoing telnet to all verified
users.  If there are problems we address 'em.  If it simply
doesn't work we go back to the way it is now.  But we would
have given it a try.

response 8 of 127:

Mar 15 03:56 UTC 1999

the crux then seems to be:

equal access for all verified users 

        vs

risking grex developing a reputation for harboring
troublemakers on the net.


Although I certainly think STeve's point is legitimate, I also think 
that if we can clarify an appropriate standard of user verification, we 
will greatly limit the potential problems Steve is concerned about.

Essentially, eliminating the anonymity will discourage the bulk of the 
troublesome elements.

The next tier of concern is grex's ability to handle the traffic flow 
increase this move would potentially represent.  As stated previously, 
offering the access to users outside grex's local phone service area 
would seem redundant, thus the most likely beneficiaries of the move 
would be relatively few in number.  Does anyone seriously suggest that 
the additional traffic from this area alone would significantly burden 
our resources?  If so, on what basis is the conclusion drawn?  Share the 
information for the benefit of the discussion.

Last is the potential loss of membership dues revenue that tying access 
to membership tips the scale just far enough to motivate.  Can we get 
any kind of estimate of what this potential is?

Let's gather the real data (i'm requesting those with the appropriate 
tools to do so, since i cannot) and revisit those elements of the 
discussion forthwith.

response 10 of 127:

Mar 15 04:51 UTC 1999

What the FSF used to do for verification of guests was to require that they
send a self-addressed, stamped envelope to the FSF, and then someone (usually
Tami Friedman) would create the account, and snail the password back.  This
made it fairly difficult to maintain anonymousness.  I think if grex wants
to allow non-members outgoing access, mailing something to such users via
US mail in some way that they have to successfully recieve the snail mail
before they get outgoing access is likely to work well.  This is a bit
difficult for grex, since the account generally already exists; I suppose
we could send mail saying that about seven days from the date we send the
mail, we will change the password to something that we include in the snail
mail, and simultaneously enable internet access for that account.

The FSF had about 1200 guest accounts, and I bet fewer than ten of them
ever caused problems.  And the few problem cases wouldn't likely cause any
more problem for the grex staff than the average eggdrop user on grex.

Most vandals want to be anonymous.  A good authentication procedure will make
them not bother to try to send the paperwork to get outgoing access on grex.

The argument about competing against ISPs is a valid one, and grex probably
doesn't want to get very many more phone lines.  I think grex ought to quite
delibrately offer a level of service on the dialups which is lower than
what ISPs offer; this can be done by keeping the dialup pool at 14.4 forever,
or not offering ppp, or whatever.  Clearly, economics make it impossible
for grex to provide the level of service an ISP does unless grex charges
everyone substantially more than the current membership dues.

If telnet is really dying out, as someone suggested, then the increase in
load this will create is irrelavent.

The choice of `Michigan' was pretty arbitrary; I'd be happy to change it to
be the case that the people who are elibigable are people who live where
grex is a local phone call.

I assume that outgoing access on grex is basically something we want to
offer to people who can't afford a $20/month ISP account; I suppose some of
them can afford a $6/month grex membership, but not all of them.  (It also
seems backwards to me, now that I think about it, that we've structured
things so that the more affluent people spend less on membership than the
poor; the poeple who pay annually pay $5/month, and the people who pay
monthly pay $6/month.  I realize that this is intended to make aruba's life
easier, but it is wrong to penalize the poor.  Also, the cost of 11 $.33 stamps
would probably be enough to motivate me to contribute annually (or at
least quarterly) anyway.

response 12 of 127:

Mar 15 05:01 UTC 1999

OK, here's the new proposed text of the motion:

    Outgoing internet access from grex for all protocols will be allowed
    to anyone in the local calling area of grex's modem pool who povides
    suitable proof of identity.  The exact definition of `suitable proof
    of identity' will be determined at the discression of the board of
    directors.

response 14 of 127:

Mar 15 12:26 UTC 1999

In response 8, a long list is laid out that never once addresses the real
burden on grex:  Adding tasks to the already too-busy staff to do list.
These tasks are grass-fire type tasks, that have to be addressed as soon as
they happen.  They are the ones that set us up for the "When is x going to
be done? -- As soon as staffer y finds some spare time"  complaint.  

Unless we have an avid volunteer, who has the time, Right Now, to take on the
new administrative burden of verification, snail mail, and follow-up on abuse,
I don't think we should make this change.

response 16 of 127:

Mar 16 14:50 UTC 1999

Item 49 is about the same subject.

response 18 of 127:

Mar 16 16:14 UTC 1999

Mixed feelings.  I kind of like the concept of non-tiered access.  I
don't think the computational/bandwidth/policing burden would be bad.  I
do think there is some administrative cost.

I like the two-way snail-mail validation process as a concept. 
Something like:
  - User sends a letter to Grex, with ID, the name of his login and
    a return address.
  - We send a letter back to the return address, containing a code
    word.
  - User runs a command "enable-inet" which prompts for the code word
    (different for each user) and turns on internet access.

Problem is, who's going to read the mail and generate the codewords and
do all that.  Yick.

response 20 of 127:

Mar 16 20:54 UTC 1999

I don't think this would put Grex in competition with ISPs, because the level
of what's being offered is very different.  Most ISP customers wouldn't know
what to do with a Unix shell account, or a non-graphical web browser.

At the same time, though, I'm not wild about putting a lot of Grex resources
towards doing a not very good job of providing a service that is commercially
available very cheaply in this area.  If I were really concerned about not
having different access levels for members and non-members, I'd be more
tempted to turn off telnet access for members than I would to turn on telnet
access for non-members.

response 22 of 127:

Mar 17 01:10 UTC 1999

I don't think it's really accurate to say "telnet is dying and the web
is taking over", or that "there is very little that telnet does that web
access will not do better".  Telnet usage isn't growing at the same rate
as web access, but that's because most people aren't using the internet
to do programming or general purpose computing, they're using the
internet to browse information.  Telnet very definitely still has its
place: if you want to do serious programming, remote system
administration, etc., you can't really do this via your web browser.
But the # of people doing serious programming and remote system
administration is not growing nearly as rapidly as the number of
non-computer type people is.

response 24 of 127:

Mar 17 04:04 UTC 1999

What janc describes in #18 seems like exactly the right idea.  Assuming
enough staff time (which I know may be a bad assumption), it ought to
be possible to write a program which whomever is processing the snail
mail can run; they enter the login name, it perhaps displays the login
name and full name, and asks for a yes/no confirmation, and then the
program picks a random code word, records the code word in a file, and
prints it on the screen.  Person writes code word on paper, and mails it off.

At the moment, my feeling is that the only problem anyone has brought up
that I have any concern about at all is the question of whether there is
a volunteer to process this mail.