• 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.

Revamp of ships' computers

Originally posted by Andy Fralix:
To continue with the line of reasoning and emply that the rules mean you un-install one program, install another and use it to perform an operation of some sort ALL in one segment of a 15 minute turn...
I understand what you're saying and I agree that the rules can be interpreted to say that. Perhaps a change in terminology would make more sense. Instead of CPU and Storage, perhaps Active and Background?
"All programs are resident in the computer. Each turn programs are selected to run in the background up to the limits listed as Storage. Each phase, programs running in the background can be moved to active up to the limits listed as CPU".

I'll be the first to admit that for large computers the whole thing becomes academic, but for smaller systems the book 2 rules would seem to give a lot more flexibility.
 
Tobias, I think we have compatable visions. I need to use a gentler tone on our minor disagreements.

Tanstaafle, the Book 2 computer rules are a superb way of simulating c. 1970 computers. I studied the Apollo flight computers and they operated much the same way.

Of course, that was before TI invented the microchip.
 
Reviewing the LBB reprint collection of books 0-8 turned up this from LBB2 page 38:
Computers
The computer installed on a ship controlls all activity within ... and conducts routine operation of all ship systems.
and from page 13:
A. The Bridge: ... basic controls, communication equipment, avionics, scanners, ...

So what can we conclude from this? CT & HG lump everything under bridge except the computer, but the bridge can't function without the computer. MT on the other hand breaks down all components and makes you select each individually.

With this in mind, why not take the simpler approach and specify that the bridge has all the components needed for all ships operations. Then specify quality of these components as civilian, commercial, or military. If you are building a private vessel that will not be carrying passingers for hire you can just have the minimum needed and save some credits. The trade off is that you have less chance of detecting and avoiding pirates and/or other potential hazzards. At the other end of the spectrum, military standard equipment has the best chances of detecting other ships and/or hazzards etc.

Now for computers: Specify size based on TL and assume that size is based on what kind of connecting technology you have at a given TL or range of TLs. Specify cost based on capacity. Go one step further and specify that ship's computers are actually triple computers for redundancy and what you get with the purchase price is 3 sets of components that comprise the unit.

I haven't had time to compute charts for costs, etc., but this method gives you the best of both worlds. Simplicity, and choice of quality of equipment.
 
Actually, simplicity still lies with me. No change to the construction rules, just cross out the word "computer" and insrt "sensors".

Obviously the bridge has an integrated data processing "backbone" to supervise ships systems, althose just as obviously these are under local control. I like the idea of layering commercial and military robustness (double and triple redundancy? 50% increase in cost, trivial increase in size). I don't think its strictly neccessary as you can't get anything less robust than "commercial" on a 30 MCr++ starship. Not and still get a clearance to enter or leave a Class A or B starport.

I think we can junk the whole concept of a "ships computer", apart from the IS buit into the bridge. It is a 1960s concept that just doesn't work well anymore.

Now there is opportunity to add richness and texture to sensors. The area of sensors increases their sensitivity and the aperature increases the resolution so size does matter, maybe more than tech level. Active sensors (radars and ladars), "safe" power for commercial, high power for military.

And scouts need gravitometers, magnetometers, and chromatographs. Also active sensors optimized for high sesolution mapping.

But that adds layers of complexity way beyond the LBB. YMMV
 
I prefer a little more maneuvering than just HG2's short and long, so sensors involve ranges, detections, locks, weapon transit times, etc. a bit more to it than just the programs.
 
Originally posted by Uncle Bob:
Replace the words ships computer with the words sensor array and no other rule change is neccessary, although you might have to fiddle the deckplans a little.
I second that. The actual computer would probably be no larger than a RL server (and even that will allow for easy access and for extensive cooling/ventilation); it will cost 5% of the total cost (and that is including the expensive interface and backups) and its volume will be subsumed in the bridge tonnage.. The rest of the tonnage/cost/EP would go to sensors and communication equipment distributed around the ship. That is, a Model/1 is 1 dton sensors/comm units that cost MCr1.9; the actual computer is subsumed into the bridge's tonnage and costs MCr0.1.

IMTUs AI's (actually pseudo-AIs as it's TL12) aren't part of the standard Computer/Sensor/Comm array, but a seperately installed LBB8 Robot Brain.

My ideas on non-ship computers:
1) Supercomputers: use full size, cost and power requirements of an equivalent ship computer, but IT ALL GOES TO THE COMPUTER (rather than sensor/comm). Double Storage by 20 and CPU space by 10. These are the central computers of large military bases, universities, big research stations etc. TL5+

2) Mainframes: use 5% of equivalent ship computer size, cost and power requirements; use equivalent ship's computer for storage and CPU space. TL6+

3) Servers: use 1% of the size, cost and power requirements of an equivalent ship computer. The Server functions as a ship's computer of one TL less than it. TL7+

4) Personal Computer: (TL7+) Cr500, 5kg; functions as a ship's computer of two TLs less than it.


5) Hand Computer: (TL8+) Cr1,000, 0.25kg; functions as a ship's computer of three TLs less than it.

6)Implant Computer: (TL9+) Cr5,000, 0.25kg; implanted in the subject's brain and connects to the lower brain's motor sections and to the sensory nerves for interface. Functions as a ship's computer of three TLs less than it.
 
Looks good, Employee, but what about upgrading the CPU, in terms of the number of programs it can handle at once?
 
Originally posted by veltyen:
Tera is Gig*1000 (Apologies purists, Yeah, I know it is an approximation).

I have multiple terabyte arrays at work. They are pretty common these days.

If you have 4 250 Gb Disks you have a terabyte available. Guess we need to come up with a new term.

Petabyte (still a rare term) is the category above terabyte. Petabyte = 1,000,000 Gigabyte (approx). Personally I don't see it making its way into common vernacular. The sizes involved are such that it isn't really important to the average user. 125,000 full length movies uncompressed takes up about a petabyte. If you watched continuously (no sleep, 24*7) this would take approx 30 years to watch. That or 2 millenia of MP3's.
Yep, and Micro$oft once that that 128k RAM was more than anyone would ever need! ;)

Sorry for the thread necromancy, but I'm reading up on overhauling Traveller computers today.
 
The earliest predictions on 'we will never need more than' were based on storing text only. Then along came b&w graphics that needed more memory. Then came color photographs that needed more memory. Then came real time sound & video that needs more memory.

How much data would it take to store a movie as a real time interactive 3D full-scale projectable holograph? I bet it takes a lot more memory.
 
Actually MS said 640KB RAM + 64KB ROM (Basic was in ROM in the original IBM PC) + Memory on the Grafics Card + BIOS, maximum of 1024KB (Since the Intel CPU couldn't do more - they should have used Motorola 68Ks
)

And you can actually build a quite decend grafics oriented system within that limits. Actually the systems where produced: ATARI ST
 
I guess that does make sense. I mean-NASA still uses crap technology from the 60's for space flight and I guess our tech then will be outdated but still in use then too. I could never see them using updated computers for anything besides astrogation anyway. (though they did use computer systems for the moon landing that didn't even possess the computing power of an old black and white ganmeboy and it didn't seem to hamper things too much) <think on that for a while-it's mind boggling!>
I think computers should be less advanced than what is available except on military aircraft/ships. It's more realistic that way.
 
Originally posted by Andy Fralix:
</font><blockquote>quote:</font><hr />Originally posted by RainOfSteel:
</font><blockquote>quote:</font><hr />Originally posted by Andy Fralix:
I have an idea for the revamp of Traveller computer design.
Post your design sequence docs, formulae, and charts, please.
</font>[/QUOTE]I haven't started yet. The idea is less than 48 hours old (at least to me) [...]
</font>[/QUOTE]Our time elapsed is now 2+ years.
 
Originally posted by Andy Fralix:
</font><blockquote>quote:</font><hr />Originally posted by RainOfSteel:
</font><blockquote>quote:</font><hr />Originally posted by Andy Fralix:
I have an idea for the revamp of Traveller computer design.
Post your design sequence docs, formulae, and charts, please.
</font>[/QUOTE]I haven't started yet. The idea is less than 48 hours old (at least to me) [...]
</font>[/QUOTE]Our time elapsed is now 2+ years.
 
Traveller and its computers....how much time is spent on this aspect of the game, I wonder.

As always, YMMV.

First, on TL. If TL5 ends in 1939 technology, then it is ludicrous in the extreme to think such a computer would be used on a TL9 starship. There were no production computers in 1939. IMTU, Computers are first available at TL6.

Second, a current server rack is about 2m high, 1m deep, and .5m wide, give or take a little. In other words about 1/4 Traveller dTon. Floor space has to be left so the components can be pulled out on their full-extension slides for maintenance. Now we've taken up 1/2 dTon. Two such racks will be 1 dTon, which is the size of a Travller Model 1.

Personally, I do not find that out of line. That is actually smaller than an IBM 4300 series mainframe from the mid-70's, which was their smallest mainframe. It is more like a VAX-sized piece of gear; a minicomputer, not a mainframe.

The CT Computer sizes are only out of line when TL increases. This problem of increasing TL only making something available, but not more effective, smaller, or less pricy is a systemic problem with CT. Jump drives, maneuver drives, etc all suffer from that problem. My solution is to reduce the size and/or cost, usually both, by 10% per suceeding TL.

Just IMTU.
 
I like the size of Traveller starship computers.

Captain: "I need a plot that will allow us to tear open the fabric of space, allow us to transition to an alternate reality, and apply an energy field to this unnatural reality that will allow us to violate the laws of physics."

Navigator: "Just a moment, captain, I will need to consult the main computer." [The navigator reaches for his PDA and finds the holder empty.] "Oops, I must have dropped it in the bar."

Steward: "That's the third one you lost this year, this time it's coming out of your pay." [Tosses a PDA to the navigator.] "Here use mine, but I'll need it back before movie time, it has all of the entertainment software loaded on it."


That is not what I think of for a ship's computer.
 
Computer technology gets smaller, but the computers themselves generally stay the same size. A modern pocket calculator might outperform ENIAC but Super computers haven't really gotten any smaller. Desktops are the same size if not bigger, ditto with laptops. I'm willing to bet Orin will fly with just as much avionics mass as Apollo, much more complicated, capable avionics, but no lighter.
 
Back
Top