|
Grex > Coop8 > #134: Should Grex get a T1 connection? | |
|
| Author |
Message |
| 25 new of 146 responses total. |
mdw
|
|
response 50 of 146:
| 
Nov 5 03:38 UTC 1996 |
One might argue with equal validity that it was fiscally irresponsible
to start grex in the first place. This situation looks no more risky
than any of the previous upgrades we've done. I don't believe there's
any risk that we would find a need to offer tiered access & lots of
perks - for one thing, the available evidence says such schemes don't
work. Also, if the gamble doesn't work out, it's certainly not the end
of the world. We could certainly cancel the T1 link if we thought it
wasn't working out. I think it's very unlikely that matters would come
to that point.
I also believe this is an opportunity that we shouldn't spend 6 months
debating. The offer we got is not going to last forever. Either we
need to decide growth is a good thing, and seize the opportunity, of we
need to decide we're too afraid to grow anymore. I very much hope the
latter doesn't happen, because that is probably the most dangerous
decision to the future of grex that we could possibly make. Deciding we
can't grow almost *certainly* means we will only come under increased
pressure to *Somehow* discriminate against non-paying members.
|
srw
|
|
response 51 of 146:
|
Nov 5 07:45 UTC 1996 |
There is a lot of sentiment on Grex against tiered access. I'm not in
favor of it either, above and beyond some minimal things like we're
doing now. I seriously doubt we'd choose that over losing our higher
speed connection. Mary's scenario doesn't ring very true to me. Much
more likely, IMO, is that we could extend the life of it for a
substantial period with a fundraiser, like we did for the disk.
I agree with Marcus that we need to decide this fairly quickly. There
is an opportunity to reduce installation costs if we order it before
12/31, and the offer from the ISP is not necessarily going to habg
around indefinitely.
I do not object to a member vote. I would welcome it. I just think that
we shouldn't wait for it if it can't be done before Christmas. I think
it probablyt can't. If we don't want to wait for a member vote, we can
get a sense for how it might go by considering the support (or
opposition) shown in this item. This has no official value, of course,
but could tell us if we're going against the tide. Ultimately I think
it is legitimate for the board to decide this, with input taken from
this item.
|
ajax
|
|
response 52 of 146:
|
Nov 5 07:46 UTC 1996 |
It's not a question of "we grow now or we're afraid to grow."
If we don't seize this opportunity, there will be other more
conservative ways to expand in the future.
|
dang
|
|
response 53 of 146:
|
Nov 5 12:36 UTC 1996 |
And this is not the same as previous hardware upgrades. Those were all "Pay
once, use forever" deals. This is a monthly obligation for 36 or 60 months,
with a fairly heafty payment if we back out. Not the same at all,
financially. However, that doesn't mean we shouldn't do it, just that we need
to be careful.
|
kerouac
|
|
response 54 of 146:
|
Nov 5 16:58 UTC 1996 |
This is about what Grex is, which is what I was pointing out
earlier. If it is a "free net" then expanding access for as many
people as possible is part of the function and should be done if
possible. If Grex is a private club set up primarily to serve its
members, then the debate is much more subjective. So what is
Grex? I've already shown that many people come here thinking this
is a freenet, through "freenet" gopher holes of one form or
another. If this is a private club, that should be made clear in
new user.
I say if grex is a type of "free net", get the T1
|
steve
|
|
response 55 of 146:
|
Nov 5 17:38 UTC 1996 |
Good grief, Mary: *We've had an anonymous donor for three years
for the current Internet link*.
Let me state that again: Grex's *current* Internet link was araranged
by an anonymous donor going through IC-Net. That been going on now
for some three years.
Yes, this is a *BIG* jump for Grex. But it isn't any bigger a jump
than the twelve people who started Grex in the first place. Or when we
made the decision to spend another $40/mo on electricty for the Eagle
disk so we could get usenet running.
I really and truely hope that Grex hasn't ossified. Hell yes, there
is "risk" involved. Most people wouldn't *dare* try and run Grex in the
first place--as an example, I recently talked to an administer of a Big
Ten college about computers, and the subject of Grex came up. This
person was flabbergasted to find out that Grex was an all volunteer
organization, and ran on less then $8,000/year. It took several minutes
of convincing to get that across--I was going to call up a board memeber
and let him talk to them! The same thing went for our Ameritech account
rep: he was pretty amazed to find out what we were, too.
Grex is damed unique. Because of this uniqueness, we've been
noticed by a major player in the net connectivity business. We're
being offered something that simply isn't going to happen every day.
If Grex fails to get this it will be more than tragic. We need it,
and we can afford it.
There is no reason not to go forward on this, from our side.
Let's look towards the future!
|
scg
|
|
response 56 of 146:
|
Nov 5 17:39 UTC 1996 |
re 53:
That depends on which previous upgrade you're talking about. Things
like buying modems, buying the Sun 4, buying disks, and things like
that, are one time costs. Adding phone lines is a longer term cost.
Grex is now already paying more for POTS lines than we would pay for the
T1.
Can we afford the T1? On our current income, no. With the promised
donations, we can certainly afford it for a year. I'm willing to assume
that if Grex's budget hasn't yet grown to the point of being able to
support it after the first year, we will have users who will be willing
to kick in some money rather than see it go away. If needed, I'll
certainly be willing to kick in something a year from now to help fund
it. Another consideration is that as more and more people get their own
Internet connections, our demand for dial-up lines is likely to go down,
potentially allowing us to shift some funds from POTS lines to the T1.
The more important question at the moment is whether we can afford not
to do this. Maybe I'm wrong here, but I'm guessing that those people
here who don't think we need a faster Internet connection haven't tried
to use the current one lately. We are now attempting to push more data
over the current connection than will fit, leading to really ugly
results like 15 second ping times. The Internet connection is often so
saturated that it couldn't reasonably be considered usable. A huge
amount of staff time is spent dealing with mail problems, many of which
are directly related to the saturation of the link. Like it or not,
Grex isn't a little local system anymore. We've got 14,000 users,
spread out all over the world. A very small percentage of our users are
still using our dial-up modems, while the rest of us (yes, even many of
us local people have to come in over the Internet) are crammed through
that tiny Internet connection. Anybody who thinks we don't need a
faster connection should try using the current one.
|
kerouac
|
|
response 57 of 146:
|
Nov 5 18:18 UTC 1996 |
get the T1...
life is short...best to live it
|
dang
|
|
response 58 of 146:
|
Nov 5 18:27 UTC 1996 |
I telnet in all the time, and the link us usually useable. Not nice, but
usable. However, I say go for it, and get the T1.
|
kaplan
|
|
response 59 of 146:
|
Nov 5 18:43 UTC 1996 |
I think #55 has convinced me, STeve. I'm in favor of a T1 link and I'll put
a tiny amount of money where my mouth is. I'll maintain my membership plus
$10 per year as long as the T1 (or better) connection is here.
|
mdw
|
|
response 60 of 146:
|
Nov 5 18:51 UTC 1996 |
The alternatives for a more "moderate" rate of growth are not nearly as
good. The two closest alternatives are a single or a double ("bonded")
ISDN link, and the prices we've been quoted for ISDN internet service
are actually quite comparable to the T1 service - but the performance we
might get is actually closer to our current PPP link. (Latency would
improve, but throughput of an uncompressed 64K ISDN link is actually
quite close to that of a 28.8K modem link which uses compression. Some
modern ISDN implementations do suport compression, but that requires
making sure both we and the vendor have compatible hardware...)
I also believe that if we can't get our act together to take advantage
of this offer, that we never will. The people who might work on
upgrading the link are likely to be disappointed, and less enthusiastic
about locating future such deals, while the people who want to see the
system stay small are likely to become more dominant in the decisions of
the system. I believe this would also mark a rather fundemental change
in the operation of the system. In the past, we've always dealt with
resource shortages on the system by growing the system's hardware. In
the absence of such, we would probably inevitably be forced to seek out
and implement increasingly draconian measures to limit usage of the
system. We've already seen a serious proposal, here, to limit e-mail
usage. On m-net, which basically *did* go with a 64K link, there is a
serious proposal to limit guests to *just* lynx (no shell access).
|
steve
|
|
response 61 of 146:
|
Nov 5 19:07 UTC 1996 |
Thanks Jeff.
|
robh
|
|
response 62 of 146:
|
Nov 5 19:46 UTC 1996 |
I worry about the risk involved with getting a T1, but I also
worry about what will happen if we don't. Generally, I just
worry. >8) I say we should give it a try, but if our donors
only want to fund it for a year, then we should only commit to
a year with Ameritech.
|
steve
|
|
response 63 of 146:
|
Nov 5 21:15 UTC 1996 |
We can do that, but we'll be looking at $333.00 per month
rather than $250.00, and we'll have to pay the $587.50 setup
charge. I'll point out that 10% of the extra two years worth
of a 3-yeard contract would be paid for, by avoiding the $587
charge.
|
scg
|
|
response 64 of 146:
|
Nov 5 23:07 UTC 1996 |
And, if I'm remembering correctly, the penalty for cancelling after one year
is having to make back payments so that Ameritech will end up having gotten
$333 per month rather than $250. If we want to plan for the possibility of
it going away after a year, we might do better to pay the extra money into
some sort of T1 doomsday fund, which could be handed over to Ameritech at the
end of the year if needed, or used to pay the first four months of the next
year if we decide we can afford to keep it.
|
ajax
|
|
response 65 of 146:
|
Nov 5 23:47 UTC 1996 |
<set long_boring_response_mode=on>
I think a lot of this hinges on affordability. Most everybody
wants a T1; hesitance or opposition is due primarily to financial
repercussions. I haven't seen a lot of number crunching online,
so here are my own estimates (which may not be accurate; please
review for errors).
Somewhat Pessimistic View
-------------------------
Here's a kind of pessimistic view of how I see this shaking out.
Suppose we get a T1 for a year, and realize toward the end that it
costs too much to continue beyond that, so we switch back to a 28.8k
link (kind of optimisticly assuming we find another donor for the
ISP charge). I'll also assume that we need a T1 CSU/DSU on our end
of the connection, and not on the other (maybe we won't need one;
I don't think that's been decided). So for this scenario, for one
year of a T1, then falling back to a 28.8k link, we've got the
following costs:
$590 T1 startup
$311/month times 12 months
$700 T1 equipment (rough estimate)
$80 to reconnect 28.8 phone lines
minus $40/month times 12 months for our 28.8k phone lines
$4622 total for one year
We've got $2400 pledged by two donors (I think that's the figure
mentioned at the board meeting, but I don't recall any figures
presented online). If so, that would leave $2222, or $185/month,
to be covered by other sources during the first year. If we didn't
get *any* new members, this would decimate our savings, but we'd
survive, if we could get back to our current 28.8k link pricing.
(Note that we'd still have T1 equipment afterwards as an asset,
which we could sell for something).
Somewhat Optimistic View
------------------------
Here's a fairly optimistic view of things. Suppose that we're
taking in enough money after one year to continue with this for
three years (after which some other deal becomes available).
I'll assume that the ISP will loan us the T1 equipment (it's not
that uncommon, as ISPs often build it into the cost of service,
and it reduces support/config headaches if they're familiar with
the equipment.) I'll also assume that we get this hooked up by
12/31, to get the T1 startup fee is waived.
$0 T1 startup
$276/month times 36 months
$0 T1 equipment
minus $40/month for our 28.8k phone lines
$8496 total for three years
Deducting $2400 pledged by two donors, that leaves $6096, or
$169/month, to be covered by other sources for a three year period.
(We'd also have two 28.8k modems freed up as assets, which could be
sold or used for other purposes).
-------------------------
If these are fairly accurate, then there's not much of a difference
in monthly costs whether we keep the link for one year or three years.
Things could get quite a bit worse than the pessimistic estimate, or a
little bit better than the optimistic view, but I think both are fairly
realistic scenarios. If so, the break-even point for finances is if
we can get and retain 30 new members fairly immediately, or more if
the member boost is delayed. (I.e., we need an *average* of around
30 more members per month, for the duration of the T1 connection).
(Extra non-member pledges/donations would help too, but I find it
useful to think in terms of members).
[Also, just for good measure, a quick estimate for a 56K leased line:
$300 (??) line startup
$165/month (M-Net's rate; not sure of contract period)
$310 CSU/DSU (What M-Net paid)
Minus $40/month for 28.8k lines
36 month total: $4750 (compared to $6096 net (w/pledges) for a T1)
This 56K leased line option all-around sucks, which is why it isn't
being seriously considered. If this were the only current option,
it would be better to hold out for discounted ISDN Internet service,
or maybe upgrade to x2-compliant modems, for an assymetric 28/56Kbps
link, when they become available next year. As an aside, M-Net went
this 56Kbps leased line route paying commercial rates to an ISP,
adding another $235/month to the the $165/month cost.]
<set long_boring_response_mode=off>
|
chelsea
|
|
response 66 of 146:
|
Nov 6 00:16 UTC 1996 |
Thanks, Rob. That was helpful.
My response back there had nothing to do with wanting
to hold Grex back or retard Grex's growth. My concern
is simply can we afford this? It's really hard to
look that one straight in the eye when you want the
link so very badly. But try. Where will we get the
money?
|
krj
|
|
response 67 of 146:
|
Nov 6 02:40 UTC 1996 |
Maybe one needs to start collecting pledges?
|
pfv
|
|
response 68 of 146:
|
Nov 6 17:15 UTC 1996 |
Just as an aside here, since it looks like all my fears are
already in print above..
Is there any evidence that this T1 is going to somehow generate
the level of donations that seem required?
I don't mean wishes and hopes: I mean statistical evidence that
the boosted telnetting populace is going to pony up $$.
If you look at M-Net 'Policy', you will see that almost no member/
patron-class users go thru the 56K.. Indeed, that some 85% of the
"net-use" (pun intended ;-) is completely guest-level.
Too, it seems to jump mail by a magnitude and lately I've been
watching a good percentage of users burning multiple ports.
|
chelsea
|
|
response 69 of 146:
|
Nov 6 17:40 UTC 1996 |
Re: #55 Good grief, STeve. ;-) An anonymous person arranging a link is
hardly the same thing as an anonymous person(s) holding a (verbal)
promissory note to Grex, to the tune of $2400, where Grex would be liable
in the event of non-payment. Again, if the donation is all up-front, no
problem.
I hate to see aspects of Grex's financial dealings kept secret
from the members. That's not what we're about.
|
steve
|
|
response 70 of 146:
|
Nov 6 19:47 UTC 1996 |
Peter, every time that Grex has taken a chance and grown its
resources to better function, more people have come forward,
and joined.
|
pfv
|
|
response 71 of 146:
|
Nov 6 21:14 UTC 1996 |
Uhhh.. Well, steve, I am sure glad that is true.
But, what you are saying is that "it'll all work out in the end"
That there is no data. *sigh* I wish you guys luck.
|
mdw
|
|
response 72 of 146:
|
Nov 6 21:57 UTC 1996 |
There are no guarantees in life.
|
ajax
|
|
response 73 of 146:
|
Nov 6 22:51 UTC 1996 |
From a member survey this summer, members estimate that they telnet in to
Grex about 30% of the time. Telnetters have a much lower membership rate
than dialers, but it's not negligible, either. There's also outbound
performance to consider: half of the members said getting outbound telnet
service was a factor in their becoming members.
There's no past data with which to predict the result of this. I entered
an item in Agora last night to ask for pledges if we get a T1, so that should
provide some indication of a lower bound of expected revenue increase.
I kind of suspect that Grex will lose a fair amount of money getting a T1.
I consider CPU speed and local dial-in accessability much more important for
increasing memberships. Though if staff decides to hold the number of ptys
constant, someone theorized that the CPU performance might improve, because
it would make mail flow more smoothly.
|
janc
|
|
response 74 of 146:
|
Nov 7 15:40 UTC 1996 |
Personally I would expect that CPU gain to be small and quickly swamped by
increased use of mail, ftp and http services.
|