|
|
| Author |
Message |
janc
|
|
RetroGrex
| 
Jul 31 12:17 UTC 2004 |
I've discovered that I have mixed feelings about moving on to the new
hardware. One of the things about makes Grex and M-Net unique is our
history. When you telnet into Grex today and run "bbs" you are in some
ways taking a trip two decades back into computer history, back to the
time when Picospan was originally designed, when a public conferencing
system was a new idea, before the Well, before the internet went public.
I think for Grex to continue to be relevant, we need to move forward.
We need newer hardware, a newer operating system, and, as I'll discuss
in another item, we need to be rethinking how conferencing works on
Grex. However, I think as we move forward there is a danger that Grex's
unique identity will disappear among thousands of web-based forums on
the Net. I think our history is worth preserving, as we move forward.
My notion is that after we have moved to the new system, the current be
kept on line, at a reduced level of service. Party and Picospan would
remain. SunOS 4.1.4 would remain. All web services would be shut down
including backtalk and webnewuser. It would no longer be possible to
send or recieve offsite email. If you want to access it, you need to
telnet or ssh in. All these services would, of course, still be
available on new Grex, but not on Retrogrex.
The reduction of services is partly to keep adminstration simple, and
partly to restore things to a state more nearly resembling the
experience that early users of Grex, M-Net or even the Well would have
seen.
The disadvantage to this plan is that it requires keeping the current
4/670 computer turned on. It's a lovely machine, sold for around
$100,000 when it was new, and built like a $100,000 product. It'll keep
working forever. It does use an unattractive amount of power, and take
rather more space than most co-lo facilities could tolerate.
Still, I think this would be a cool thing to do.
|
| 67 responses total. |
cross
|
|
response 1 of 67:
|
Jul 31 15:37 UTC 2004 |
Cool, but impractical for the reasons you mention (power, space, etc).
One doesn't need an old Sun to preserve computer history. The users
are the thing that counts.
|
prp
|
|
response 2 of 67:
|
Jul 31 19:52 UTC 2004 |
Is there a VM-Sun program?
|
drew
|
|
response 3 of 67:
|
Jul 31 20:25 UTC 2004 |
That's what I was thinking (emulation).
|
janc
|
|
response 4 of 67:
|
Jul 31 20:50 UTC 2004 |
Actually, I think you do need an old machine to give the whole thing
deep legitimacy. Knowing you are using a genuine old computer, not a
modern machine pretending to be an old computer makes a world of
difference, and the difference will grow bigger with time. A modern
machine emulating an old machine would just be a shallow fraud.
In the 4/670 we have a somewhat unique item. When I mentioned this at
the staff meeting, Steve Andre' raised the idea of really pushing
backward, and bringing up a Sun 2 or Sun 3 or one of the older Sun 4
machines that we still have pieces for. I want nothing to do with that.
I remember those machines. They were flakey as all get out, failing
regularly for mysterious reasons. The 4/670 it their direct
technological decendent, running the same software, using the same VME
bus, but it's a rock solid machine, in many ways better than anything
that can be bought today. I see no reason it couldn't run forever with
minimal effort. This makes it nice - an undoubted dinosaur, but a hardy
one. The perfect base for a project like this. Retro-cool enough to
give the idea heft, but stable enough not to make it a pain in the ass.
I recognize that I'm one of the few Grexers ever to have taken this
machine apart, and it's hard to convey what a lovely creature it is
without doing that. I guess to a lot of people, it would be mysterious
even if you have done that.
For the time being, I don't think power or space are such big factors.
I think it is much more important to the future of Grex to figure out
what we mean to the future of the internet, and I think an homage to the
past is a slightly paradoxical, but effective way to achieve a part of
that. It nails down something about Grex's historical identity in a
substantive way.
And as Dan said: "Cool but impractical..."
Why, oh why would we turn down a chance to do something cool just so we
can be practical? If we start making practicality our goal, Grex is
lost.
|
janc
|
|
response 5 of 67:
|
Jul 31 21:00 UTC 2004 |
Remember, when people donate money to Grex, they want to feel like they
are supporting something unique, not just another cookie-cutter web
forum. The coolness of having an old machine like this stay on the net
will increase with time. You'll be able to say to people "send us a few
dollars to help keep this old power hog turned on". I'd not care to
make any guarantees, but I think there is a chance that that could turn
into a positive asset that actually brings in more than it costs. There
may be a day when the geeks at a co-lo facility would leap at the chance
of hosting an real live working Sun 4/670 on their site, just for the
variety from all those identical rack-mounted boxes. It's like the
people who keep antique cars on the road, restore old buildings, or
re-enact civil war battles. Obsolete does not equal trash. There is
value in this, and maybe even financial value.
|
jp2
|
|
response 6 of 67:
|
Jul 31 21:42 UTC 2004 |
This response has been erased.
|
cross
|
|
response 7 of 67:
|
Aug 1 00:18 UTC 2004 |
When you consider that timesharing under a "real" version of Unix
is becoming more and more unique, I see little difference between
SunOS 4.1.4 on VME based SPARC and OpenBSD whatever on x86. To me,
the thing that makes grex grex is the shell prompt, not the fact
that it's a rack of old hardware in a little room somewhere.
Is the SPARC machine cool? Sure. I dare say it isn't as reliable
as Jan makes it out to be: there have been problems with memory
boards, multi-processor configurations, etc through the years.
Strolling back through Garage has given me the impression that
while yes, it is stable, it's not *that* stable. I certainly think
a modern commodity machine is going to be *more* reliable,
particularly as time goes by.
|
twenex
|
|
response 8 of 67:
|
Aug 1 02:04 UTC 2004 |
It's a great idea. Old computers are getting harder and harder to find. I
refuse to get rid of my old Amigas and Spectrum, and remain in despair of ever
getting hold of a VAX (I shudder to think of the cost and potential risks
involved in shipping one over the Atlantic; they seem pretty thin on the
ground here in the UK or continental Europe.)
To a certain extent, I agree that you need people to keep history alive; but
what happens when the people have moved on?
|
cross
|
|
response 9 of 67:
|
Aug 1 02:27 UTC 2004 |
If the people have moved on, the community is dead. ``This is my favorite
hammer; I've replaced the handle 17 times, and the head 23 times to keep
it usable!''
Here's a suggestion: Donate the old machine to a computer museum with the
priviso that it must be kept running as long as feasible and accessible
as a BBS system in the style of grex. Make sure it has a copy of
Picospan and newuser installed, but most everything else turned off.
Then it becomes someone else's problem, but the historical value of the
machine/software combination can be preserved.
Personally, I'm not terribly stokked on the idea of subsidizing grex
as a computer museum when other such things exist and I'm really into
subsidizing grex as an interesting socio-technical experiment.
|
twenex
|
|
response 10 of 67:
|
Aug 1 03:20 UTC 2004 |
I like that suggestion, too. Perhaps the members ought to vote on some
proposals?
|
janc
|
|
response 11 of 67:
|
Aug 1 13:44 UTC 2004 |
Do you actually believe that there is any computer museum who wants
Grex's not-terribly-rare computer enough that they'd except a donation
that included a condition requiring them to do something forever? I
think that's more in the realm of the impossible than the impractical.
Do other such things exist? Oh, I know there are computer museums, but
had the impression they turned their computers off and certainly did not
put them on the net and open them to the public.
As far as space goes, I don't think that retroGrex would necessary
always have to be in the same place as grex. If Grex were to move to a
co-lo that didn't want to house a sun 4/670, it could probably move to
someone's basement - it doesn't need dialin lines or a fast network
connection.
My hope would be that a continuing conferencing community would grow up
on retroGrex, probably different in character to the one on Grex. It
would be small and geeky. It might appeal especially to old M-Net/Well
users. It might work out pretty well - a certain level of barriers to
enter can sometimes encourage growth of a good community. But the
barriers may be too high, and the levels of interest might be too low,
and usage might fade to zero. If that happened I'd advocate turning it
off.
We'd have web pages on Grex that told about retroGrex, gave instructions
for how to connect to it (telnet or ssh1), and how to use bbs and party.
(I might be tempted to roll back to an older version of party.)
|
cross
|
|
response 12 of 67:
|
Aug 1 15:07 UTC 2004 |
Note that I said as long as practical. I don't think they'd be interested
in only the hardware, but I *do* think they'd be interested in the
hardware/software combination and the preserving the community aspect.
Personally, if I were at a computer museum and someone told me about
something like this, I'd think it'd be a pretty cool piece of the net
to keep going.
I'm sorry, but I just don't see a `retrogrex' as being in keeping with
grex's mission or charter. If someone wants to put the old Sun 4/670
in their basement at the end of a small DSL line or whatever, then more
power to them and I certainly have no problem with a web page on grex
pointing them to it, but I really don't see it being feasible to keep
it going with scarse volunteer time that could be going to grex proper.
To summarize my position, I don't have a problem with the old grex
machine continuing to run, just as long as it's done outside the context
of grex itself.
But this is really all academic at this point. The new machine isn't
even running the latest version of the operating system, nor is it on
anywhere near ready to accept the user load from grex itself. Let's
worry about getting that online before we worry about what to do with
the old hardware.
|
twenex
|
|
response 13 of 67:
|
Aug 1 15:20 UTC 2004 |
Keeping a server open for telnet connections at all is pretty retro in this
day and age.
|
ric
|
|
response 14 of 67:
|
Aug 1 15:32 UTC 2004 |
I don't think, personally, that the average user cares WHAT Grex is running
on. I certainly don't. I find it, at times, to be obnoxiously slow. The
people here are what makes Grex unique, not the hardware it runs on.
|
mfp
|
|
response 15 of 67:
|
Aug 1 21:11 UTC 2004 |
Yeah, Rick, but you just say that 'cause you use it for interacting with
people.
ANYWAY< GUYS< I THINK WE"RE FORGETTING ONE THING> WE HAVE TO GET NEW GREX
RUNNING FIRST!
AND ALSO JAN"S IDEA ABOUT CASTRATING OLD GREX (BY USING AND OLD VERSION OF
PARTY AND ALL< WTF?!) IS JUST AS MUCH OF A "CHEAP FRAUD" AS RYUNNINNG SOME
EMULATOR PROGRAM.
|
ric
|
|
response 16 of 67:
|
Aug 2 00:55 UTC 2004 |
re 15 - the ones that come here to interact with other people are the ones
that are most important to the Grex community.
|
gelinas
|
|
response 17 of 67:
|
Aug 2 01:36 UTC 2004 |
I think the 'community-building' aspect of retroGrex is well within the
purpose of Cyberspace Communications, Inc. I also think the maintenance of
retroGrex will become easier, once FTP is turned off.
I would like to see us do both.
It should be noted that the discussion of retroGrex *is* relevant to the
migration to the new machine: The existence of retroGrex affects the software
set that will run on the new machine. The new machine does not depend upon
retroGrex, though.
|
tod
|
|
response 18 of 67:
|
Aug 2 15:33 UTC 2004 |
re #11
Can't Cyberspace add Museum to its non-profit list of deeds?
|
albaugh
|
|
response 19 of 67:
|
Aug 2 17:56 UTC 2004 |
I would object to any "rat packing" scheme that cost more money than necessary
(unless those who wanted it were willing to foot the bill). Jan's #5 is *not*
why I am a member of grex.
|
cross
|
|
response 20 of 67:
|
Aug 2 18:57 UTC 2004 |
Regarding #17; Why does the existence of a retrogrex affect what set
of software the new machine will run in any way, shape, or form? And
how will a retrogrex help build a community any more than the current
current configuration (which is essentially retrogrex) does? If it
will, why are we bothering with nextgrex at all? We should sell the
machine for 50c on the dollar and use the money to pay for something
else.
|
janc
|
|
response 21 of 67:
|
Aug 3 14:28 UTC 2004 |
I don't think the existance of retrogrex effects the set of software on
Grex in any direct way.
Personally I think that Grex right now is on a slow downward slide. I
think we are seeing that in our income and in the overall quality of the
conferences. I suspect (though I haven't checked) that we are seeing a
decline in pure volume of activity in the conferences. When I look at
Grex's future, I see us sliding along in the trail of M-Net. That
doesn't mean Grex is doomed, but it isn't exactly an exciting future.
There is already an M-Net. Who needs two?
I think it's time to try making some changes on Grex. Moving to a new
machine is an important step in facilitating that, because it widens our
options, but in itself it solves nothing. We need to be prepared to
make deeper and more experimental changes to the system. My confidence
that anyone can be found to do that work isn't high, but I hope we can.
This is not something I am entirely enthusiastic about. A slow slide
into senility is one possible "death" of Grex, but getting revamped and
revised into something different is also a sort of "death". We need to
do everything we can to preserve the ideals and style of Grex as it
evolves, but it's an uncertain processes, and there is always a strong
constituency for doing nothing. I think having retroGrex around will
help people (including myself) who are nervous about changes feel a bit
more confident that the baby isn't going to be lost with the bath water,
and to preserve a snapshot of Grex's history, so people who come along
can have a bit more context to think about Grex's future.
So I think if we do this, and if it works (that is, the system maintains
a decent critical mass of users) then it might have an indirect
psychological effect on the evolution of Grex. But I would still want
to keep as many old Grex elements as possible on new Grex too -
including shell access and a picospan compatible command line interface.
|
mfp
|
|
response 22 of 67:
|
Aug 3 18:40 UTC 2004 |
WHY NOT MERGE WITH M_NET?!
|
gelinas
|
|
response 23 of 67:
|
Aug 3 19:28 UTC 2004 |
By keeping Picospan on the current machine, we eliminate (or at least
reduce) the demand for Picospan on the new machine. Which leaves us free to
design new conferencing software that better meets our needs (and desires).
I don't think News, with any interface, meets our needs.
I'm not sure that RetroGrex will "build a community," but I think it will
help us preserve at least some of our current community.
NB: RetroGrex _is_, essentially, currentGrex. No new development is
required to keep it going.
|
cross
|
|
response 24 of 67:
|
Aug 4 00:09 UTC 2004 |
True, but the amount of effort required to keep it going is effort
that could be used on the new machne, but won't be.
Why don't you think News with a custom interface will foot the bill,
Joe? I think of it as a message store and convenient way to share
conference data, not as an extension of the pre-existing USENET
infrastructure. I contend that a clone of the picospan user interface
backed up by an NNTP server will look exactly the same as `real'
picospan to the end user.
|