response 50 of 142:

Jan 20 04:37 UTC 1995

I only have a guess to back this up, but I think that the 10% covers a
large part of outstate Michigan which happens to have a local connection
to Merit, and perhaps fewer other affordable net options.  This represents
a large area that will always be beat down in terms of relative total
numbers. 

response 52 of 142:

Jan 20 07:38 UTC 1995

Right:  Out-of-state is a different situation, but I wrote outstate.  I
meant the out of metro areas in Michigan with relatively less population
and probably far fewer service choices.  Sault Ste. Marie, Alpena,
Marquette & many other Merit sites.

  I use the merit "public" gateway to grex frequently.  I can afford to
seek an alternative, and think I will have something to choose from.  I
think the question for grex is if something like 4 or 6 merit ports would
make any sense as far as serving the people of Michigan.  Would this be
convenient for a substantial group of people?  Would it create a funding
interest great enough to underwrite the expense through memberships and
donations?  Or are there enough widely distributed services available
*through which grex will keep an active profile* that the effort would be
essentially redundant.  

  Grex could afford to plunk for the setup and annual fee, and then have a
"support-a-port" membership drive each year.  Money donated could be
collected to determine that X ports would be funded for the coming year. 
By the calculation above a $250 donation would support 1 port for one
year.  20 $62.50 donations would support 5 ports. If there was too little
interest, things could then go by the wayside.  I'd say about 4 ports
would be a working model.  A worthwhile exposure for the money needed to
setup, and 5 of 50 would be %10, seemingly very close to the actual usage
for this route, even though it is "free" now. 

  This is Merit's pound of flesh.  Is there a better way?

response 54 of 142:

Jan 20 09:24 UTC 1995

I'm not sure I entirely understand the "month" part of jep's "$250 per
"average user per month" per year statement, in 46.  If it's the number
of simultaneous users, why not say that? If it's the number of connect
hours, then a connect month makes sense, but suggests we could be
talking $250/mo not per year - or twelve times jep's figures.  I've
heard a rumour elsewhere that it costs Merit $60/mo to maintain one
dial-in port, which suggests that $250/yr would not cover all of Merit's
costs.

Two additional issues that Jep glosses over are "connection costs", and
"authorization server".  I presume the "connect costs" are the costs of
the actual link between grex & merit; which could be divided into
hardware grex would have to buy, charges to ameritech or whomever for
the wire to get to merit, and whatever charges merit would impose on top
of that for the hardware at their end.  I am also not clear how much
flexibility we might have here.  For instance, would merit care if we
implemented a packet radio link?

The "authorization server" issue might be the most sticky.  Not
necessarily technically (they already support kerberos), but the real
issue would be if there is any per-user verification or paperwork
requirement on top of that.  The question there is if we can continue to
run newuser as we do now, or if we'd be required to validate users after
they ran "newuser" before they could actually use grex.

response 56 of 142:

Jan 21 11:59 UTC 1995

The nice thing about Merit, obviously, is its wide reach.  After we get 
the move behind us:

1. We should try to calculate our "average load."
2. Get some new figures from Merit
3. See if those users that currently use the Merit dialins would be will-
   ing to support that cost.

We've got quite a few users that get to Grex this way so I don't think
it's out of the question.

response 58 of 142:

Jan 21 18:17 UTC 1995

   What statistics are you looking for when you say average load?

response 60 of 142:

Jan 21 18:57 UTC 1995

   Have you tried perhaps taking a single class at lssu in order to
get access that way?
   (I know, drift--sorry).

response 62 of 142:

Jan 21 23:00 UTC 1995

I suppose when the university based computer systems stop serving the
general public, they should also restrict university accounts to
university business:  classwork, research, and verifiable academic
affairs. (including the content of e-mail).  It's like being able to take
university vehicles across country for vacations and drinking binges.  If
anybody can do it, then exactly how academics use the resources is beside
the point, as they can be "public" too.  If only a select few can have
subsidized accounts, then the use becomes suspect. 

response 64 of 142:

Jan 22 01:09 UTC 1995

re 62:
        Through my job at the University of Michigan, I get a lot more
computer and network access than the general public has ever gotten from
the University, and for the most part I don't use it for anything having
to do with my job, nor do I think there's anything wrong with letting
University employees but not the general public have access to it for
recreational purposes (I mean, it would be nice for the general public to
have lots of access to university computers, but I don't see a moral
problem with the current situation).  I regard it all as a perk of the
job, and there is no more wrong with the University giving computer
related perks to its employees but not the general public than there is
with it giving pay checks to its employees but not to the general public. 
Or is somebody going to argue that I shouldn't be spending my pay check
for anything non work related?

response 66 of 142:

Jan 22 02:53 UTC 1995

I'll have to admit that I've never met anyone that was getting any work
related perks that had any trouble rationalizing their existence.  I hope
the Highway Department never uses the model used in so many cases by our
educational institutions with public resources.

  I think it's fine that students and employees have liberal access to
computer resources.  As to the public, let them eat cake, or Delphi. 
The internet is a rather tangled mess, and perhaps someday a niceperson
genius like Bill Gates will come along and rescue us all via our TV sets.

response 68 of 142:

Jan 22 03:46 UTC 1995

That's OK if it is announced. But its not. So, no one knows if it
is "OK" or not. In fact, you'll probably be told before long that it
isn't (the same thing happened with telephone use, where the employee
has to figure out what personal calls are "OK" and which not). 

response 70 of 142:

Jan 22 05:35 UTC 1995

   Rane, the things like mosaic and news, X.500 and the other things 
included in the "core" services are so basic to net related things that
it doesn't make much sense to even try and meter them.  As amazing at
(as) the UM is in charging for things, someone drew the line somewhere
and said in effect that its harder to meter these things than just have
a flat rate fee for them, and be done with it.
   It is interesting to note that ITD want to be able to sell these
services to the general public. I still don't know what the rates will
be, but in theory I'm in a queue of people waiting to find out.

response 72 of 142:

Jan 23 05:43 UTC 1995

Why aren't you a board candidate too, tnt, with such good ideas? B-P

response 74 of 142:

Jan 23 14:45 UTC 1995

I thought I had access to merit because I pay taxes to the state of Mich.