response 126 of 164:

Aug 13 17:38 UTC 1997

You are bending the twig until it breaks. Our practice of allowing telnet
only to members is a great membership builder. Consider what would happen
if we gave it away like e-mail. We are already, for complex reasons,
giving perqs to members so you can't go off the deep end like that. Yes,
limiting telnet reduces the load on our link. Limiting e-mail would also,
even if just delaying some to slack periods. It may increase membership
too. 

response 128 of 164:

Aug 14 02:22 UTC 1997

Good questions all. We (and you) have often stated that if we did not
limit outgoing telnet, we would be bombed with users. I take that to be a
judgement that it is a desired feature for which, in part, members join.
Nobody "commonly become active participating members in the conferences",
but some do. Members might be more likely to. Less e-mail traffic would
improve conferencing too. All benefits of a slight bending of the e-mail
twig. Of course ISDN makes some of these things moot - but will force
upgrades of system and memory, etc., mostly for e-mail. The argument that
we can absorb enormous worldwide e-mail traffic if we only had an ISDN
link is wearing blinders. Is that why we are installing an ISDN link? Will
it really help, if people find the bottleneck in the e-mail pipeline
removed? I doubt it. 

response 130 of 164:

Aug 14 04:52 UTC 1997

Alternatively, we could conclude that we have detered a very large number
of people that would use the free outgoing telnet, leaving Grex with a
higher fraction of members than we would otherwise have. One has to look
at the leverages. Allowing non-member outbound telnet would produce many
more users than new members. Allowing non-member outbound e-mail produces
many more users than new members. We eliminate the former and we cut down
hugely on users but only a little on members. But the system is still
slow, in part, because of the e-mail load. So, cut back a *little* on
non-member e-mail, and we would cut down some on e-mail, but again only a
little on members. Except, the improved performance of the system for
*conferencing* would counteract that.  These arguments are all theory, of
course. But what do members have against experimenting with different
stratagems? Its not as though making changes is irreversible. 

response 132 of 164:

Aug 14 07:39 UTC 1997

We could also experiment with cutting off the left hand of offenders who
overflow their mailbox, or translating all the web pages on grex into
French.  Either or both might make a positive change in the system.

Making a change in itself costs resources.  (There are obviously
tremendous technical obstacles to implementing manual amputation, even
though practical working examples exist in Saudia Arabia.) If one is to
avoid just making random flailing mutations in the hopes of improving
the creature, it's desirable to have some notion of goals and priorities
("just what is better?") and to use some sort of self-consistent model
of the universe to try to make the best incremental improvements
possible ("don't march downhill to go uphill, if there's a better way to
get there.") Remember, in nature, most mutations are either neutral or
negative; and lots are rather immediately fatal.

In this case, we have plenty of evidence that many people consider
e-mail to be "pretty important".  One consistent model that explains
this is that many people here think of grex as a "destination" or
"home", instead of just a "thoroughfare" to get to other places.  In
this model, "e-mail" is an "essential" home service, along with unix
access, picospan, and all the other "standard" attributes of an account
on grex.  We even have testimonial evidence that e-mail brings some
people into the rest of grex, in particular, the conferencing, which has
already been identified as an important goal of grex.  Also, in this
model, outgoing telnet access is clearly seen as not a good use of grex.
Using this model; restricting e-mail looks bad.  It's diminishing
several core values of grex, for instance, "grex as a home" and "grex is
a place to communicate (read and write) with other people." This doesn't
prove that restricting e-mail on grex is a bad idea, of course.  But it
does suggest that, if this is an experiment we are willing to
contemplate, then we should have some good reason to discount this model
of grex.  Perhaps we aren't interested in "grex is a home", or "grex is
a way to share information with friends"; perhaps we don't like the
people here we've attracted in the past with e-mail; and clearly, we
should be willing to make changes that will definitely piss off a
distinct majority of the vocal participating conference members that
we've seen responses from right here in this item.

response 134 of 164:

Aug 14 20:11 UTC 1997

Lopping off hands is like lopping off twigs, a bit more brutal than the
bending I am suggesting. 

I think we like all the things you suggest we aren't interested in
(perhaps). 

Yes, people think Grex is a home for e-mail. That's what we are talking
about.  Do we want to make Grex a home primarily for somewhat low-end
e-mailing?  That sounds more like a house than a home.  If not, the thing
to do is diminish its utility and increase the utility of what we do want
Grex to be a home for.

I don't see that we have to "discount" any model of grex to make
experimental changes to test procedures that might improve the quality of
grex. A "home"  is not just a free handout. In fact, there are a lot of
users here that argue that free handouts are counterproductive. Why
doesn't that apply to Grex?  Remember, I am not suggesting going any way
at all toward a 'for profit' operation, but just suggesting that users
should participate in exchange for their benefits in the home. If they
refuse to participate...well, does that sound like a home? 

response 136 of 164:

Aug 15 00:44 UTC 1997

   But they have a lot more ptys than we do, too.  If we had 254 ptys
open here (the max or maybe 255) then we might be able to handle all
the load.  But we can't, so I think it is realistic to say that we're
going to be swamped.

response 138 of 164:

Aug 15 16:18 UTC 1997

Nether.net has an ethernet speed connection to the internet, unlike grex.

response 140 of 164:

Aug 17 02:05 UTC 1997

different mailers - as in pine, elm, etcetera?

response 142 of 164:

Aug 18 11:21 UTC 1997

four minutes after #131, #132 does an excellent description of
'what it is to be grex.'  however, if 132 comments on 131, there is
absolutely zero reason/rationale or argument 'to discount this model
of grex.'  and the metaphor of a 'home' is, quite properly, dead on.
  
every 'model' however has a 'scale.' every home has *individual* 
capacities within each component that is assembled into taht 'model
home.'
  
there are so many of these, so many of those, a bunch of other things,
this many rooms, that many toilets, some quantity of couches &/or chairs,
some extra mattresses, has or doesn't have a dishwasher / laundry / 
garbage disposal ... (the laundry might even be coin-operated!) ...
and some limit on inbound electrical capacity ........
  
so if these resources are able to be distributed on a fairly
equitable basis, on demand, except for ... uhhhhhh ... water pressure
then no one gets enough water, or pressure regardless of the fact that
every other resource is successfully distributable.
  
of courese not enough water pressure also means the toilets dont
flush too well, the shower doesn't, the laundry cant, the garbage
disposal overheats, the kitchen sink overflows with unwashed
dishes, the grass dies, flowers cant get watered, the vegetable
garden is a pipe dream and the damn cars are filthy.
  
of course the 'model house' can bed down 14,000 in the attic and
60+ live ones in teh living room, on the patio, inthe basement bar,
at a couple of dinner tables, and a whole group onthe dying lawn.
  
and then all 60+ need to use the single toilet.  oops.
  
the single toilet and the unavailable water pressure are reasonably
distributable, say, to 40+ live ones....but 60+ ? ummmmmm, something
inthe 'model house' would seem to need adjustement before the
fan is the substitute for the toilet.

response 144 of 164:

Aug 20 15:58 UTC 1997

I like the direction this item is taking! Plainly my motion in #0
is technically unworkable.  But I would indeed like to see some
"twig benders".  I don't think Grex should be a "home" for low-end
e-mailers.  As scg said in #129:

If we allowed outbound telnet for free, we would be swamped.

Right now we allow unlimited e-mail for free and we are swamped.
Let's continue looking for ways to drain the swamp.

response 146 of 164:

Aug 20 17:08 UTC 1997

As long as it applied across the board (members and non-members alike) I 
don't see a problem with that idea -- but I'm still not convinced it's 
necessary.

I think Jan's mail machine idea is the best one so far.