embracing retro 'puters

[BwapTED]

I was looking through The Imperial Fringe.

Library searches on world info are measured in hours.

I've often argued that the computer tonnages aren't really that huge when one considers what a ship's computer has to do and all the shielding, cooling, redundant features, etc.

But let's back up.

Traveler computers are retro. Traveller info tech is retro. That's the baseline.

Why not just embrace that?

Libraries contain micro-books of the sort so common in sci fi of the early and mid 20th Century. Search engine? You mean the micronized card catalog?


Jump cassettes? Loads of cassettes!

TL 5-7 computers resemble computers from Babbage Engines through 1960s machines, but after that it gets funky.


This aint our future.

edit-

Looks as if I misread something in my skimming of that adventure. It's surveys, really, not library data searches.
Ooops!

 

[ShawnDriscoll]

I was looking through The Imperial Fringe.

Library searches on world info are measured in hours.

I've often argued that the computer tonnages aren't really that huge when one considers what a ship's computer has to do and all the shielding, cooling, redundant features, etc.

But let's back up.

Traveler computers are retro. Traveller info tech is retro. That's the baseline.

Why not just embrace that?

Libraries contain micro-books of the sort so common in sci fi of the early and mid 20th Century. Search engine? You mean the micronized card catalog?


Jump cassettes? Loads of cassettes!

TL 5-7 computers resemble computers from Babbage Engines through 1960s machines, but after that it gets funky.


This aint our future.

I've always treated the Traveller Universe as an alternate future. I get more players starting games quickly that way, instead of playing Retcon the Universe at the table.

 

[Ulsyus]

...
Library searches on world info are measured in hours.

...Traveler computers are retro. Traveller info tech is retro. That's the baseline.

Why not just embrace that?

Libraries contain micro-books of the sort so common in sci fi of the early and mid 20th Century. Search engine? You mean the micronized card catalog?


Jump cassettes? Loads of cassettes!

TL 5-7 computers resemble computers from Babbage Engines through 1960s machines, but after that it gets funky.

This aint our future.

Well, it was CT so what was included was internally consistent with other CT products. There's no reason it all couldn't be played using T5, which would provide a bit more scalability WRT technology.

If the PCs were in a TL6 or 7 system then perhaps the searches would take hours.

 

[mike wightman]

I didn't have a problem with the size of CT computers back in the day and I still don't

A Traveller ship computer, even in CT, days if anything is too small not too big.

It is equivalent to today's supercomputers, not an I-pad.

What a model 1 can do:

run a nuclear fusion reactor (ever seen the size of the server rooms at CERN?)

run the environmental systems - this includes gravity and acceleration compensation

run the avionics, sensors, comms

control a maneuver drive

run or plot an n-body hyperdimensional transit

Now harden the thing so exposure to radiation in space isn't going to cause it to go belly up...

I see this argument a lot - mostly from people who think a desktop can do all this stuff. They can't. Our desktops also have one other advantage, the interwebs which is actually several millions tons of computer equipment (a local network can be smaller ).

As to a library data search taking hours - there is a difference between Wikipedia and actual research. I would allow players to get a brief overview in a couple of minutes, but proper info takes longer.

 

[whulorigan]

A Traveller ship computer, even in CT days if anything is too small not too big. It is equivalent to today's supercomputers, not an I-pad.

What a model 1 can do:

• run a nuclear fusion reactor (ever seen the size of the server rooms at CERN?)
• run the environmental systems - this includes gravity and acceleration compensation
• run the avionics, sensors, comms
• control a maneuver drive
• run or plot an n-body hyperdimensional transit

Now harden the thing so exposure to radiation in space isn't going to cause it to go belly up.... . .

And, just to be clear, can do almost all of these things simultaneously. The CT Hand Computer was supposedly the equivalent of a Model/1, and it is TL11, not TL7-8, and is thus definitely much more powerful than an I-pad.

Plus, I like the interpretation (I believe Supplement Four on CotI has previously said it) that the size of the CT Ship's computer necessarily also includes the volume of the physical hardware of all of the ships electronics (Comms, Sensors, possibly back-up/redundant computer processors, etc - since CT has no design sequence for these obviously necessary but otherwise taken for granted pieces of shipboard equipment) which when things like antennae and detectors are considered, are not small devices.

The size of the ship's computer only starts to become an issue when you add in additional design elements to the standard design rules that detail the electronics systems separately.

 

[RandyB]

And, just to be clear, can do almost all of these things simultaneously. The CT Hand Computer was supposedly the equivalent of a Model/1, and it is TL11, not TL7-8, and is thus definitely much more powerful than an I-pad.

Plus, I like the interpretation (I believe Supplement Four on CotI has previously said it) that the size of the CT Ship's computer necessarily also includes the volume of the physical hardware of all of the ships electronics (Comms, Sensors, possibly back-up/redundant computer processors, etc - since CT has no design sequence for these obviously necessary but otherwise taken for granted pieces of shipboard equipment) which when things like antennae and detectors are considered, are not small devices.

The size of the ship's computer only starts to become an issue when you add in additional design elements to the standard design rules that detail the electronics systems separately.

As with any of the games from that era, when you try to go beyond the level of abstraction presented and zoom in on greater detail, it stops "making sense". 'The fault lies not in the games, but in ourselves.'

 

[BRJN]

The section in Interstellar Wars about Vilani computers (which are hard-wired for ONE task and not reprogrammable) would also help explain why the ship's computer is so much bulkier than your iPad.
One Vilani world is TL6. Despite being "technologically backwards" (and really do use building-sized central computers) they still could sustain the Cold War, equip James Bond, and create an Apollo Program from made-in-homeworld materials.

 

[mike wightman]

I assume any Traveller TL7+ computer system to be as capable as what we have today.

Siri, Cortana, Alexa - they are all voice command interfaces. Imagine how smart they will be at TL8+

Now how big is a TL9 quantum computer needed to operate a jump drive?

I think the disconnect in Traveller computers is that the underlying technology and capability improves with TL - the size increase models this but we have the usual problem that the CT rules for building lower TL stuff at higher TLs are missing.


A TL15 model 1 computer is going to be very different to a TL 5 model 1 computer, IMHO the TL5 model 1 should be a lot bigger than 1t.

 

[Enoki]

Well, it could be they have to be hardened for the use they'll see. That could mean a considerable increase in size to allow them to work in vibration, radiation, extremes of temperature, etc.

If a ship were to say lose its atmosphere and the temperature drop to near absolute zero, most commercial computers today would no longer work. That's one example.

 

[whartung]

But let's back up.

Traveler computers are retro. Traveller info tech is retro. That's the baseline.

Why not just embrace that?

Because it's not Jules Verne, and it's not Steampunk.

We've seen where physics takes us as knowledge of electronics progresses. We've seen computability as a mathematical concept when made real in machine, both digital and analog.

Nothing we have today in terms of electronics and computing is revolutionary. Even the transistor wasn't revolutionary. It's all been incremental refinement over the years. There's no new physics here.

But saying we don't have computers, we have EtherCalculators, or whatever, that's as handwavium as the M-Drive.

Finally, the computers in Traveller, from a game play and mechanic perspective, are not that important. Space combat is space combat, whether the computer is one ton or 20..

 

[BwapTED]

Because it's not Jules Verne, and it's not Steampunk.

We've seen where physics takes us as knowledge of electronics progresses. We've seen computability as a mathematical concept when made real in machine, both digital and analog.

Nothing we have today in terms of electronics and computing is revolutionary. Even the transistor wasn't revolutionary. It's all been incremental refinement over the years. There's no new physics here.

But saying we don't have computers, we have EtherCalculators, or whatever, that's as handwavium as the M-Drive.

Finally, the computers in Traveller, from a game play and mechanic perspective, are not that important. Space combat is space combat, whether the computer is one ton or 20..

No, not steampunk. That's pretty recent, arguably starting with The Difference Engine.

You are closer with Verne, but off by a few decades.

I am referring to computers and information tech as they appear in (a lot of) science fiction written between the 1920s and the 1960s.

Foundation , by Asimov, and Space Viking, by Piper, provide good examples.

 

[BwapTED]

I see this argument a lot - mostly from people who think a desktop can do all this stuff.

But you have never seen it from me.:badger:

On the contrary, I have made all the same points you've made about why star-ship computers should be big. I noted that in the first post.

It's the combination of sizes, the way programs work in CT, and the apparent slowness of searches in the library that suggests an alternative take on computers and information technology, to me.

2-3 hours to look up one element of a planet's UWP seems pretty slow.

 

[BwapTED]

I just got a tip about the searches. Looks like I missed something important.

The slow searches aren't what I had thought.

I still might embrace early/mid 20th Century sci fi computers and information tech for the fun factor, but it was a misread bit that launched me on this tangent.

-----------------


But, on the subject of 'retro' computers, I do note that Book 3, 1977, lists the Model 1 as TL 5 and the Model 1 bis as TL 6.

 

[kilemall]

I always enjoyed the thought of stranding players on a TL5 world and having to get an analog/gear computer built and programmed to get their ship going again.


As far as RL time constraints, I ran general ledger runs that took 7 hours back in the 80s.


I now deal in a more distant manner with GL runs that take- 7 hours.


Of course the amount of capacity between a 1988 and a 2018 computer is exponential, but so are all the shortcuts, compilers that assume vast quantities of resources, less tight programming in general that reduces dev costs, and the raw amount of data, regulations to apply, sophisticated reporting, etc.


So I wouldn't worry too much about run long times. Just have big enough differences between TL/Model results that it 'feels' like there is value in going high end, look at the complexity of the task as a major variable, and do what is right for the game.


I feel a formula coming on....


C⁴/TLxM= minutes


where C is complexity, TL is Tech Level of computer and M is model of computer.


Complexity is derived by ballpark estimates of what is wanted.


So retrieving a wiki type Library Data is C1 and thus 1 minute divided by virtually everything into near instant, audio request for information Siri-style is C2 and 16 minutes, correlation of a 100 square km area survey data into reporting and a database is C5 and thus 625, etc.


Isn't really exponential like we have experienced in our lifetimes, but good enough for players to notice between running something through their Model/1 TL 9 clanker, a nice Model/4 TL12 science station system, and the Model/9 TL15 flagship.


I'd probably factor in abstract values for what the computer packages are given it's role- a science computer might crunch that survey through fast but not do well analyzing the tactical intentions of any enemy, while the Navy computer couldn't summarize a biome to save it's digital life.

 

[JimMarn]

While I can see my old Amiga A500 computer, at 7.14 MHz CPU 32 bit internal, 16 bit on the buss, taking a long time to search oh a terabyte of data being in the hours. I don't see a more recent computer with a 3 GHz clock and SSD drives taking that long.

This slowness, and able to only handle a very few, or only one, computer program at a time, has always bothered me at CT. Computers now aren't that slow.

As for big and bulky computers, I have seen a Cray YMP-2 search through a terabyte of tapes, put the one the software is looking for into the tape drive, pull up a data set of ocean currents, and run software on it.

That Cray had multiple software queues. So it ran several programs at the same time. And this is late 1980s technology.

That computer was used to model tornadoes, hurricanes, and ocean currents. That could take hours.

But just getting a page or two of info, I don't see how that could take hours. Every large computer I have been an operator on, did such things rather quickly.

On the opposite side, one of the graduate Computer Science school professors gave one of his classes a oject to come up with a program, this was under MS-DOS 4.x, that could be run by a small ship captain, like a shrimp boat, to predict where the hurricane was going.

He was using Turbo Pascal. Then he tried Turbo C. There were problems. 1) the software simply couldn't handle enough variables, 2) they kept finding things that simply couldn't be added in. Ocean bottom, as the storm approaches North America, had an effect.

The software typically got the same track as actuallity, for less than a day... a few times it was correct for 24 hours of a prediction. Then it would wildly go off somewhere... occasionaly going where hurricanes have never gone, like heading towards the Antarctic, or Africa, and bounce around all over and make landfall in Australia.

So he gave it up and had his grad students do a differnet project.

I have occasionaly wondered if today's desktops could manage it or not.

 

[Lycanorukke]

I have occasionaly wondered if today's desktops could manage it or not.

A major catch with computers today is not the CPU, but memory speed. While the CPU can chug away at billions (or more) of calculations a second, the memory transfer is an order of magnitude slower. So a lot of the time the CPU is just spinning its wheels while waiting for the memory to catch up. Even expensive super fast memory is still slower than the processor.

There are ways to mitigate this such as caching and predictive processing (where the CPU 'guesses' what you will do next and pre-emptively calls the memory) but they are limited to a degree.

But speaking of supercomputers, the US has just commissioned Summit at 200 petaflops.

 

[BwapTED]

Arguing against my own retro concept, but....


About jump cassettes, could they be some evolution of the newer sorts of cassettes used now for information storage?

 

[whartung]

About jump cassettes, could they be some evolution of the newer sorts of cassettes used now for information storage?

Cassette is a form factor, doesn't necessarily mean there's a tape media in it.

Similarly, "tape" can just be the vernacular from the old days like "dialing the phone" or why we have floppy disk icons on modern computers.

 

[BwapTED]

Cassette is a form factor, doesn't necessarily mean there's a tape media in it.

Similarly, "tape" can just be the vernacular from the old days like "dialing the phone" or why we have floppy disk icons on modern computers.

True, though I do like the idea of a tape.

 

[nobby-w]

[ . . . ]
run a nuclear fusion reactor (ever seen the size of the server rooms at CERN?)

The big server farms at CERN are for crunching sensor data and running simulations and other applications. Anything involved in controlling the accelerators are large running RTOS platforms.

The rest is computationally trivial up to

[ . . . ]
run or plot an n-body hyperdimensional transit

Getting warm - this is more fun. Perhaps you need a cryogenically cooled quantum device to do this, although canonically you can buy tapes with a particular route on them.

[ . . . ]
Now harden the thing so exposure to radiation in space isn't going to cause it to go belly up...

I see this argument a lot - mostly from people who think a desktop can do all this stuff. They can't. Our desktops also have one other advantage, the interwebs which is actually several millions tons of computer equipment (a local network can be smaller ).

I think folks underestimate just how powerful today's commodity hardware really is. A Really powerful™ computers don't have to be very large at all. A modern GPU has throughput measured in trillions of calculations per second.

However, for a CT level ship, a computer is a complete avionics package. It has to be quad redundant, hardened, radiation shielded, fitted with auxiliary power supplies, shielded conduits for cabling. It's not just a computer as well, it's sensor systems, antennae, comms and many other things. You don't have to say how powerful it needs to be. It's got all sorts of other hardware other than just the CPU.