• Welcome to the new COTI server. We've moved the Citizens to a new server. Please let us know in the COTI Website issue forum if you find any problems.
  • We, the systems administration staff, apologize for this unexpected outage of the boards. We have resolved the root cause of the problem and there should be no further disruptions.

Computer usage

Computer rules - a long-time complaint is that they are not realistic. People look at a laptop and say why a system in the tons.

For that, consider this:

  • a starship computer has to be in a shock-mount frame - obviously!
  • It needs a UPS - obviously!
  • It also needs access space for repairs - perhaps not so obvious.
So, here are typical rack mounts: http://www.rackmountsolutions.net/Server_Rack_Perforated.asp. let's assume 1 door, so the floor space is the specified width by the width added to depth; for the first 36" model shown the floor space is about .67m x 1.67m or roughly a half dton.

Now the other thing, is that the published stats for a Traveller Model 1 are NOT representative of the 1970's. It is, however, for late 50's second generation computers. Here are some references for the IBM System/360 - the first Third generation computer, i.e. first to integrated circuits instead of discreet components.
https://www-03.ibm.com/ibm/history/exhibits/mainframe/mainframe_PR360.html
http://www.beagle-ears.com/lars/engineer/comphist/c20-1684/fig100.jpg

Note the price in the first link - $133,000 in 1964 for the base model.

In the second link, note the size - you can compare by the desk there. And that is a Model 30 - Traveller equivalent would be a Model 2. Roughly a half-ton.

The easiest way to fix the computers is to assume the stats are for TL5, then for every TL after that, halve the size and halve the cost. Then, a TL7 Model 1 would be 1/4 dton and cost Cr250,000 while a TL9 Model 1 would be 1/16 dton and cost Cr65,000. The cost is still high, but the size is about right.
 
I highly disagree about your 1970s comment Bill, and wonder how much work you have done with mainframes to make that assertion.

Our 1980s mainframes certainly did take up that sort of room.

And the conception of program capacity matched the way machines worked then with partitions. 'Multiprogramming' was a radical thing that first got done in the big machines, and eventually trickled down to PCs.

I pretty much forgive Traveller it's computer sins, and find it much more fun to justify the system retrofitting then mocking it.

After all, the seminal future computing/hacking RPGs, GURPS Cyberpunk and Cyberpunk 2013, specced systems that sounded wildly powerful at the time but ultimately got it wrong too.
 
I highly disagree about your 1970s comment Bill, and wonder how much work you have done with mainframes to make that assertion.

Our 1980s mainframes certainly did take up that sort of room.
I did assembler on the Navy's AN/UYK-7 in the late 70's.

Did Cobol, Pascal, and Basic on MVS-8000, a mini, in the early 80's.

Did Cobol and Basic on Dec System-10 and System-20 in the mid 80's.

Did Cobol and PL/1 with IMS on a succession of IBM mainframes starting in 1986 with 4341, 4361, 4381, 3083, 3092, ES390, and Z390 up through 2005.

Did you ever look inside those 4300 series? They were half empty - if a CFO paid out a quarter mill, they wanted to see something BIG.

What Models did you work on?
 
Did you ever look inside those 4300 series? They were half empty - if a CFO paid out a quarter mill, they wanted to see something BIG.

I remember replacing a 4300 series (which was roughly the size of a washer and dryer combo) with, effectively, a modern tower sized computer (it was a Data General 88K machine). I'll never forget the power phallus that it had jutting out of the wall.

I also remember replacing an ancient, room filling, over head towering NCR main frame that some how managed to survive the extinction level asteroid strike, with a Sun pizza box SPARCstation…and a 25 port Terminal concentrator.

I don't think the band printer that NCR used would fit in the back of my truck.

I would have liked to spend a couple years on an IBM OS/390 system and/or and AS/400. Doing work, in anger, making a business run, back before the IBM machines were just Linux containers. Never got the chance, I think being exposed to that way of thinking would have only helped broaden my base. Those machine did things Different.
 
:rofl:

I remember a cub scout group through the hospital's computer room, and one of them pulled the big red switch on the front of the 4381 - that was the Production Hospital system. Down for 4 hours or so...

The kid thought it was funny as hell.
 
I did assembler on the Navy's AN/UYK-7 in the late 70's.

Did Cobol, Pascal, and Basic on MVS-8000, a mini, in the early 80's.

Did Cobol and Basic on Dec System-10 and System-20 in the mid 80's.

Did Cobol and PL/1 with IMS on a succession of IBM mainframes starting in 1986 with 4341, 4361, 4381, 3083, 3092, ES390, and Z390 up through 2005.

Did you ever look inside those 4300 series? They were half empty - if a CFO paid out a quarter mill, they wanted to see something BIG.

What Models did you work on?

I would consider your naval computer to be a Model/1.

The first one I was on was an NCR Criterion, like I said Model/1 bis or maybe even Model/2.

Then we switched to Burroughs.

B7800.

B7900.

Both of these filled the 'new' computer room, especially when one considers 'storage'.

The computer room, don't recall the exact dimensions, but back of the napkin figuring a 2 foot square tile floor system works out to 2x3 tiles per dton. Guessing something like '32-40 tons'- not counting external cooling towers and power/UPS mains.

A15.

That was when serious miniaturization started taking place- an I/O-datacomm cabinet became Cisco router sized, and disks started coming in arrays. We also had room for a DEC VAX, DEC Alpha, ES9000 and an increasing number of servers.

The ES9000 was a baby machine relatively speaking, only used for financial projection/budgeting.

About 1995 we dumped Burroughs and started using IBM for the main system, which by that time was smaller, I'd say the storage arrays and robotic (later virtual) tape systems were larger- then we merged and ended mainframe usage in that data center in 2001 or so.

That gets us out of the 80s, when yes your game model would match what happens. But what's the fun in that?

Besides, I wouldn't put most mainframes in space that were not specifically avionic machines (like the Saturn V machine).

If your big IBM machines were smaller then the B-series, keep in mind they were air-cooled, as I recall a lot of IBM machines were water cooled and you need to factor that into your space usage.
 
:rofl:

I remember a cub scout group through the hospital's computer room, and one of them pulled the big red switch on the front of the 4381 - that was the Production Hospital system. Down for 4 hours or so...

The kid thought it was funny as hell.

We had a programmer that liked doing tours, and would let his hand hover over the master computer room emergency power off button for dramatic purposes, and one time he meant to just point it out and hit it.

On the second time he did that (swearing he never hit it) we put a tape label box over it, naming it after him and wrote on the cover the legend of why the box was there. He never went near it again, we could still tear off the cover and hit the power, never had to.
 
Back
Top