Back to the main page

Mailing List Logs for ShadowRN

Message no. 1
From: Sascha Pabst <Sascha.Pabst@****.INFORMATIK.UNI-OLDENBURG.DE>
Subject: Mainframes
Date: Thu, 6 Jul 1995 15:40:54 +0200
Hi...

Just a question:
Are there any rules/prices/whatever on the (open???) market that
cover Mainframes and ICs? What does a (small/medium/large) coorporation
need in 'puters? How much can they afford I may see from their
profile, but _how_do_I_calculate_the_costs_???

Sascha
--
+---___---------+-----------------------------------------+------------------+
| / / _______ | Jhary-a-Conel aka Sascha Pabst | The one does not |
| / /_/ ____/ |Sascha.Pabst@**********.Uni-Oldenburg.de |learn from history|
| \___ __/ | or | is bound to live |
|==== \_/ ======| Westerstr. 20 / 26121 Oldenburg | through it again.|
|LOGOUT FASCISM!| *Wearing hats is just a way of live* | |
+---------------+-----------------------------------------+------------------+
Message no. 2
From: "J.D. Falk" <jdfalk@************.ORG>
Subject: Re: Mainframes
Date: Thu, 6 Jul 1995 14:28:26 -0400
On Thu, 6 Jul 1995, Sascha Pabst wrote:

> Just a question:
> Are there any rules/prices/whatever on the (open???) market that
> cover Mainframes and ICs? What does a (small/medium/large) coorporation
> need in 'puters? How much can they afford I may see from their
> profile, but _how_do_I_calculate_the_costs_???

Mainframes are all but outdated even today -- for example, I'm
currently typing this while using a Sun SparcStation 20 as an xwindows
terminal. This Sun, which costs under US$6000, is more powerful than most
mainframes of even five years ago, and fits nicely on my boss's desk.
FASA hasn't explained the hardware end of the Matrix all that
well, but my guess (based mostly on the new matrix rules) is that each
"node" is equivalent to what we'd now consider a seperate computer (for
you Unix dweebs, switching between nodes is just a fancy rlogin.) What
FASA calls an SPU is a more powerful system, designed to correlate data
from the various nodes connected to it; FASA's CPU is the /central/
processor, correlating data from the SPUs and various nodes.
Please note that this is not official FASA doctrine, only my own
opinions -- but it makes sense, kinda, sorta, right?

---------========== J.D. Falk <jdfalk@************.org> =========---------
| "Fifteen years ago 'cyberspace' referred to the file system |
| on a CDC6600 running NOS." |
| -- James D. Murray (he was joking) |
----========== http://www.cybernothing.org/jdfalk/home.html ==========----
Message no. 3
From: Sascha Pabst <Sascha.Pabst@****.INFORMATIK.UNI-OLDENBURG.DE>
Subject: Re: Mainframes
Date: Thu, 6 Jul 1995 21:12:08 +0200
>
> On Thu, 6 Jul 1995, Sascha Pabst wrote:
> > Just a question:

J.D. Falk answered:
> Mainframes are [...description of physical components...]
Great, I thought so, too. But maybe I didn't make that clear: I need
prices for a computer system one can build in SR-Universe, and in fact
it does not care _what_ it is, just: What need I, what does it cost, and
-maybe- is it illegal?
Sascha
--
+---___---------+-----------------------------------------+------------------+
| / / _______ | Jhary-a-Conel aka Sascha Pabst | The one does not |
| / /_/ ____/ |Sascha.Pabst@**********.Uni-Oldenburg.de |learn from history|
| \___ __/ | or | is bound to live |
|==== \_/ ======| Westerstr. 20 / 26121 Oldenburg | through it again.|
|LOGOUT FASCISM!| *Wearing hats is just a way of live* | |
+---------------+-----------------------------------------+------------------+
Message no. 4
From: "S.F. Eley" <gt6877c@*****.GATECH.EDU>
Subject: Re: Mainframes
Date: Thu, 6 Jul 1995 16:37:30 -0400
> Mainframes are all but outdated even today -- for example, I'm
> currently typing this while using a Sun SparcStation 20 as an xwindows
> terminal. This Sun, which costs under US$6000, is more powerful than most
> mainframes of even five years ago, and fits nicely on my boss's desk.

Hmmmm. I disagree with you, but first: Where the <HADES> did you get a
SPARC 20 for only $6,000? Was that used, or educational slash-down? I've
been pricing Suns for a prospective Internet Service Provider, and SPARC
20's start at $17,000 on the low end and move up from there. I'm not saying
you're wrong, I'm just wondering where you got such a deal.. The price you
describe is about right for a good SPARC 5 on Sun's end-user price list.


> FASA hasn't explained the hardware end of the Matrix all that
> well, but my guess (based mostly on the new matrix rules) is that each
> "node" is equivalent to what we'd now consider a seperate computer (for
> you Unix dweebs, switching between nodes is just a fancy rlogin.)

Sure, distributed computing makes sense, but saying "mainframes are all but
outdated" is a total misconception. The state-of-the-art moves up on the
high end just as much as it moves on the low end. For instance, your
SPARCStation 20 may be well and good for almost any work you might do as
an individual, or even for a mid-sized workgroup, but for a major campus
network, corporate OR academic, you'd need a bigger machine. My college
(Georgia Tech) currently uses a SPARCCenter 2000, about a half-million
dollar server, distributed with three SPARC 1000's as the student and
faculty computer system. More specialized research gets done on KSR
massively parallel system, SGI Reality Engines, or a Cray YMP. Whether
or not you call these machines "mainframes," they're still big, powerful,
and hideously expensive. And necessary for some forms of work.

In the Shadowrun world, the size and complexity of a node is probably an
indicator of the size and complexity of the machine running it, and that
will vary as widely as corporate needs and resources will allow. Yes, it
seems logical that each node is its own machine; but there's no reason to
believe that a "Sun 20" workstation-class system will be adequate for all
nodes. Joe's Data Shack could probably get away with the 2050's equivalent
of a Sun 20 for its CPU, and smaller or more specialized machines for its
SPU's, slave nodes, and datastores. (Read "file servers" in today's terms.)
A larger and more complex Matrix system like, say, Mitsuhama would have much
more sophisticated machines at many nodes, and Mitsuhama's own
top-of-the-line at the CPU.

(*subject change*)
Rules to handle the budgeting: Hmmmmm, that's a VERY good question. Mostly
it's a matter of plot apprehension -- what Matrix does the corporation need
to challenge the players? Okay, they've got that one. >8-> If you're
planning to build a specific corporation into a campaign in greater detail,
I can see why you'd be interested. If there aren't rules, there should be.
Perhaps you should make some up yourself.. Just set some base prices and
multipliers for Security Levels and system node ratings, add some overhead
for networking, and tweak it 'til it makes sense to you.

If you do this, post your work to the list! Several of us would be
interested.


Blessings,

_TNX._

--
Stephen F. Eley (-) gt6877c@*****.gatech.edu )-( Student Pagan Community
http://wc62.residence.gatech.edu| "The forces of evil are
My opinions are my opinions. | persistent, sir."
Please don't blame anyone else. | -- _Under a Killing Moon_
Message no. 5
From: "J.D. Falk" <jdfalk@************.ORG>
Subject: Re: Mainframes
Date: Thu, 6 Jul 1995 17:45:21 -0400
On Thu, 6 Jul 1995, S.F. Eley wrote:

> Hmmmm. I disagree with you, but first: Where the <HADES> did you get a
> SPARC 20 for only $6,000? Was that used, or educational slash-down? I've
> been pricing Suns for a prospective Internet Service Provider, and SPARC
> 20's start at $17,000 on the low end and move up from there. I'm not saying
> you're wrong, I'm just wondering where you got such a deal.. The price you
> describe is about right for a good SPARC 5 on Sun's end-user price list.

I guess we get good deals...either that, or I didn't hear my boss
right. If you want, EMail me privately, and I'll see if I can find out
who our contact is.

> Sure, distributed computing makes sense, but saying "mainframes are all but
> outdated" is a total misconception. The state-of-the-art moves up on the
> high end just as much as it moves on the low end. For instance, your
> SPARCStation 20 may be well and good for almost any work you might do as
> an individual, or even for a mid-sized workgroup, but for a major campus
> network, corporate OR academic, you'd need a bigger machine. My college
> (Georgia Tech) currently uses a SPARCCenter 2000, about a half-million
> dollar server, distributed with three SPARC 1000's as the student and
> faculty computer system.

We've got over 2,000 users, not to mention a whole lot of
dedicated lines, on three BSDi servers with Pentium/90 chips. The system
load average is rarely as high as 2. Of course, we've also got a lot of
router power.

> More specialized research gets done on KSR
> massively parallel system, SGI Reality Engines, or a Cray YMP. Whether
> or not you call these machines "mainframes," they're still big, powerful,
> and hideously expensive. And necessary for some forms of work.

But how processor-intensive is the kind of stuff most companies
will be doing? Transfers of data storage, EMail, accounting....
You're right, though, real serious work takes real serious
processor power. I guess it all depends on what type of system it is, and
how rich they are. *grin*

---------========== J.D. Falk <jdfalk@************.org> =========---------
| "It involves such a marked departure from normal human conduct |
| that we can't help being drawn to it." -Edward L. Greenspan |
----========== http://www.cybernothing.org/jdfalk/home.html ==========----
Message no. 6
From: Forgotten Horror <phinar@**.CENCOM.NET>
Subject: Re: Mainframes
Date: Thu, 6 Jul 1995 23:07:14 EDT
Ignoring all the doubletalk about what kind of systems they are, what
kind of power they have, etc, I think a nice ballpark estimate could
be made using the deck prices. Figure the node's MPCP at (# of
successes required) * (node rating) and build from there. This
assumes, of course, that the corp is planning on building their own
stuff. If not (if PC's want to establish a matrix presence, but gots
no decker to cook their chips) you could price the work out based on
stock decks.

Probably, based on node function, you would want either a flat
multiplier (the easy way out) or a feature chart (which could start to
be *really unpleasant*). When I first started playing the game, I
thought I might like to design the sourcebook that covered this,
largely because my characters were forming a small security corp.

I have, with much experience and a bit of disgruntlement with decking
in general, decided that I was mistaken about how intriguing that
could be, and I am now hoping like Sisyphus that VR2 will address this
concern...


<G>
(phinar@******.net)
Message no. 7
From: Robert Watkins <bob@**.NTU.EDU.AU>
Subject: Re: Mainframes
Date: Fri, 7 Jul 1995 14:52:38 +0930
J.D. Falk wrote:
>
> On Thu, 6 Jul 1995, Sascha Pabst wrote:
>
> > Just a question:
> > Are there any rules/prices/whatever on the (open???) market that
> > cover Mainframes and ICs? What does a (small/medium/large) coorporation
> > need in 'puters? How much can they afford I may see from their
> > profile, but _how_do_I_calculate_the_costs_???
>
> Mainframes are all but outdated even today -- for example, I'm
> currently typing this while using a Sun SparcStation 20 as an xwindows
> terminal. This Sun, which costs under US$6000, is more powerful than most
> mainframes of even five years ago, and fits nicely on my boss's desk.

Mainframes are still around in 2050+... I remember one module (I _think_ it
was the one that came with the First Ed GM Screen) that had a mainframe in
it. :)

> FASA hasn't explained the hardware end of the Matrix all that
> well, but my guess (based mostly on the new matrix rules) is that each
> "node" is equivalent to what we'd now consider a seperate computer (for
> you Unix dweebs, switching between nodes is just a fancy rlogin.) What
> FASA calls an SPU is a more powerful system, designed to correlate data
> from the various nodes connected to it; FASA's CPU is the /central/
> processor, correlating data from the SPUs and various nodes.
> Please note that this is not official FASA doctrine, only my own
> opinions -- but it makes sense, kinda, sorta, right?

This pretty much meshes with what I believe exists in the Shadowrun
computers... assymetricall multiprocessor machines, with processor modules
(they're assymetric as the processors are probably specialized), with the
CPU being the "controlling", or master, processor. (This even fits in
nicely with the system design rules in VR)

In this context, a mainframe is just a computer with a VERY powerful
control processor, and some other (probably LOTS) of very powerful slave
processors. While not strictly necessary, they're probably all in the same
box to reduce communication overheads.

--
Robert Watkins bob@**.ntu.edu.au
Real Programmers never work 9 to 5. If any real programmers
are around at 9 am, it's because they were up all night.
*** Finger me for my geek code ***
Message no. 8
From: Jani Fikouras <feanor@**********.UNI-BREMEN.DE>
Subject: Re: Mainframes
Date: Fri, 7 Jul 1995 15:08:48 +0200
> This pretty much meshes with what I believe exists in the Shadowrun
> computers... assymetricall multiprocessor machines, with processor modules
> (they're assymetric as the processors are probably specialized), with the
> CPU being the "controlling", or master, processor. (This even fits in
> nicely with the system design rules in VR)
>
> In this context, a mainframe is just a computer with a VERY powerful
> control processor, and some other (probably LOTS) of very powerful slave
> processors. While not strictly necessary, they're probably all in the same
> box to reduce communication overheads.

This makes sence from a technicall point of view, but the problem is that
such a computer will not be economically viable. Why buy such a powerfull
and complex (meaning expensive) system when you can solve your problems
by using a much cheaper and adequately powerfull smaller machine.
I am not saying that the smaller machine will be just as powerfull
as the parallel monster you described, I am just saying that its going
to be more cost-effective.
But this definitely isnt SR :)

--
GCS d H s+: !g p1 !au a- w+ v-(?) C++++ UA++$S++L++$>++++ L++>+++ E--- N+ h*(+)
W(+)(---) M-- !V(--) -po+(---) Y+ t++ 5++ R+++ tv b++ e+ u++(-) f+ r- n!(-) y?

Moderator of alt.c00ld00z (coolness in general)

Further Reading

If you enjoyed reading about Mainframes, you may also be interested in:

Disclaimer

These messages were posted a long time ago on a mailing list far, far away. The copyright to their contents probably lies with the original authors of the individual messages, but since they were published in an electronic forum that anyone could subscribe to, and the logs were available to subscribers and most likely non-subscribers as well, it's felt that re-publishing them here is a kind of public service.