CT+?

[Uncle Bob]

Thank God Sigg seems to agree with me. The simplest solution is to ignore computers (until they get AI) and use the dtons for sensors. Junk the programming and use the Book 5 to-hit tables.

T4 is an attempt to temporize. It gives a more modern "feel" but it is still orders of magnitude away from reality. It might have seemed a good solution in 1980.

Use Book 2 or T4 if you want to condemn Traveller to being laughed at by another generation of tech-savvy people, people who have even a vague idea of what a TL7 computer can really do. Me, I got tired of apologizing for CT about 1979.

You want to find gas giants? The data processing is almost trivial, but you need really good sensors.

 

[Uncle Bob]

Thank God Sigg seems to agree with me. The simplest solution is to ignore computers (until they get AI) and use the dtons for sensors. Junk the programming and use the Book 5 to-hit tables.

T4 is an attempt to temporize. It gives a more modern "feel" but it is still orders of magnitude away from reality. It might have seemed a good solution in 1980.

Use Book 2 or T4 if you want to condemn Traveller to being laughed at by another generation of tech-savvy people, people who have even a vague idea of what a TL7 computer can really do. Me, I got tired of apologizing for CT about 1979.

You want to find gas giants? The data processing is almost trivial, but you need really good sensors.

 

[Chuck Anumia]

Originally posted by robject:
Do you suppose there ought to be a program for detecting gas giants in nearby, uncharted systems? How big would this program be?

Actually I believe this is a matter for Better sensors and arrays than computer programs (unless the programs are being used to filter the information being fed into them by the sensors?).

 

[Chuck Anumia]

Originally posted by robject:
Do you suppose there ought to be a program for detecting gas giants in nearby, uncharted systems? How big would this program be?

Actually I believe this is a matter for Better sensors and arrays than computer programs (unless the programs are being used to filter the information being fed into them by the sensors?).

 

[robject]

Thanks, Chuck.

And Bob, the reason I don't laugh at CT computer rules is because Traveller really isn't about reality for me.

On the other hand, I can't remember the last time I've actually used the CT computer rules...

 

[robject]

Thanks, Chuck.

And Bob, the reason I don't laugh at CT computer rules is because Traveller really isn't about reality for me.

On the other hand, I can't remember the last time I've actually used the CT computer rules...

 

[far-trader]

Originally posted by Uncle Bob:
Thank God Sigg seems to agree with me.

I'm not looking for a fight but have a question for you Uncle Bob. How much space total would be needed for a computer built today on the order of something that can:

Handle a small intranet of say 10-20 interfaces and process different realtime 3D graphics for each simultaneously.

Accept input without error from even marginally trained users.

Monitor and balance a gigawatt powerplant.

Correlate and coordinate multiple sensor and communication channels.

Monitor people passively through biometric observation to evaluate threat potential in a wide range of situations.

Monitor and balance the life support of several people.

Monitor and balance the status of several critical care level patients.

Manage tactical combat across light seconds.

And all of it 24/7 for a century or more with minimal maintenance, often without any maintenance except annually, with never a failure. No blue screen of death, nor even a file not found. No corruption of files, no loss of data, and any file in the computer you want is retrieved instantly.

Please include the cooling needed and reasonable access for the rarely needed maintenance.

Also make it immune to changes in atmospheric pressure from vacuum to at least a full atmosphere, and not subject to humidity changes.

And it has to be built hardened and redundant enough to take damage from a near missile or laser blast and still perform without interruption.

Do that and you have a CT model/1 except for the real-space and jump-space avionics and navigation requirements. The first is negligible, the second we don't have a real clue. That's why I chose to ignore them in the requirements.

Let's put your budget in the neighbourhood of US$ 6 million.

 

[far-trader]

Originally posted by Uncle Bob:
Thank God Sigg seems to agree with me.

I'm not looking for a fight but have a question for you Uncle Bob. How much space total would be needed for a computer built today on the order of something that can:

Handle a small intranet of say 10-20 interfaces and process different realtime 3D graphics for each simultaneously.

Accept input without error from even marginally trained users.

Monitor and balance a gigawatt powerplant.

Correlate and coordinate multiple sensor and communication channels.

Monitor people passively through biometric observation to evaluate threat potential in a wide range of situations.

Monitor and balance the life support of several people.

Monitor and balance the status of several critical care level patients.

Manage tactical combat across light seconds.

And all of it 24/7 for a century or more with minimal maintenance, often without any maintenance except annually, with never a failure. No blue screen of death, nor even a file not found. No corruption of files, no loss of data, and any file in the computer you want is retrieved instantly.

Please include the cooling needed and reasonable access for the rarely needed maintenance.

Also make it immune to changes in atmospheric pressure from vacuum to at least a full atmosphere, and not subject to humidity changes.

And it has to be built hardened and redundant enough to take damage from a near missile or laser blast and still perform without interruption.

Do that and you have a CT model/1 except for the real-space and jump-space avionics and navigation requirements. The first is negligible, the second we don't have a real clue. That's why I chose to ignore them in the requirements.

Let's put your budget in the neighbourhood of US$ 6 million.

 

[mike wightman]

The reason I listed the computer tonnage as negligable in the earlier post is that I think it can be safely "hidden" within the bridge tonnage.

The computer's complexity and ability to run more demanding software should increase as TL increases, but, as Jame has put it in the past, a TL9 model 1 computer should be a lot smaller than the TL5 model 1.

I wonder if the TL column on the computer table needs a revision?

If the model 1 became TL 7, with bis and fib models just costing more money rather than requiring a higher TL as well, then the TL progression remains the same once you get above the model 2.

 

[mike wightman]

The reason I listed the computer tonnage as negligable in the earlier post is that I think it can be safely "hidden" within the bridge tonnage.

The computer's complexity and ability to run more demanding software should increase as TL increases, but, as Jame has put it in the past, a TL9 model 1 computer should be a lot smaller than the TL5 model 1.

I wonder if the TL column on the computer table needs a revision?

If the model 1 became TL 7, with bis and fib models just costing more money rather than requiring a higher TL as well, then the TL progression remains the same once you get above the model 2.

 

[robject]

So, T5 (ah, this isn't a T5 forum, but anyway) probably won't count computers as having their own tonnage.

Heck, I've lobbied for a reduced bridge size.

Sigg, the more you twiddle with CT computers' size + cost per TL, the closer you get to T4.

 

[robject]

So, T5 (ah, this isn't a T5 forum, but anyway) probably won't count computers as having their own tonnage.

Heck, I've lobbied for a reduced bridge size.

Sigg, the more you twiddle with CT computers' size + cost per TL, the closer you get to T4.

 

[Uncle Bob]

Originally posted by far-trader:
How much space total would be needed for a computer built today on the order of something that can:

Handle a small intranet of say 10-20 interfaces and process different realtime 3D graphics for each simultaneously.

Not the job of a CT "ships computer". But a couple of routers, say a small shoebox, and each wklrkstation will need a accellerated graphics card.

Accept input without error from even marginally trained users.

Again, a CT computer is run by professionaals, even if not by programmers. But a palm unit running Windows CE can take input from users who are barely literate.

Monitor and balance a gigawatt powerplant.

Manifestly, not something a CT computer is expected to do. Simple proof: when the ship's computer is destroyed the power plant neither shuts down nor blows up. OTOH, as part of tbe power plant percentage down in Engineering is an enormous number of processors, maybe even something like "Deep Blue" dedicated to plasma phyics.

Correlate and coordinate multiple sensor and communication channels.

Not something the Ship's Computer is expected to do: should be part of the sensor package/bridge. Depending on how closely you want these correlated it could take anything from the size of a pack of ciggaretts to a .5mx.5mx1.5m rack.

Monitor people passively through biometric observation to evaluate threat potential in a wide range of situations.

Depends more on your internal sensor package, Easily handled by a desktop computer.

Monitor and balance the life support of several people.

Not a CT job. The LSS doesn't go TU if the computer is destroyed. But local processors equivalent to a GHz laptop will be needed.

Monitor and balance the status of several critical care level patients.

Not a function of a CT ships computer. Requires a local processor in the hospital bed.

Manage tactical combat across light seconds.

This will take some power. Say a shoebox sized processor.

And all of it 24/7 for a century or more with minimal maintenance, often without any maintenance except annually, with never a failure. No blue screen of death, nor even a file not found. No corruption of files, no loss of data, and any file in the computer you want is retrieved instantly.

So they are all running Unix. Otherwise you can build in redundany pretty easily.

Please include the cooling needed and reasonable access for the rarely needed mainenance.

Using discrete processors allows them to be widely seperated for cooling and maintenance.

Also make it immune to changes in atmospheric pressure from vacuum to at least a full atmosphere, and not subject to humidity changes.

And it has to be built hardened and redundant enough to take damage from a near missile or laser blast and still perform without interruption.

Trivial with integrated circuits and solid-state construction

Do that and you have a CT model/1 except for the real-space and jump-space avionics and navigation requirements. The first is negligible, the second we don't have a real clue. That's why I chose to ignore them in the requirements.

Actually, you are demanding a great deal more than CT requires from a model 1. Remember a model 1 cannot do many of the things you asked about at all and the rest it cannot do simultaneously. All it can do at one time is normal space navigation, which you admit is trivial. That program has to be unloaded before the computer can be used for anything else.

Let's put your budget in the neighbourhood of US$ million.
More like $60K. Might be a lot less

 

[Uncle Bob]

Originally posted by far-trader:
How much space total would be needed for a computer built today on the order of something that can:

Handle a small intranet of say 10-20 interfaces and process different realtime 3D graphics for each simultaneously.

Not the job of a CT "ships computer". But a couple of routers, say a small shoebox, and each wklrkstation will need a accellerated graphics card.

Accept input without error from even marginally trained users.

Again, a CT computer is run by professionaals, even if not by programmers. But a palm unit running Windows CE can take input from users who are barely literate.

Monitor and balance a gigawatt powerplant.

Manifestly, not something a CT computer is expected to do. Simple proof: when the ship's computer is destroyed the power plant neither shuts down nor blows up. OTOH, as part of tbe power plant percentage down in Engineering is an enormous number of processors, maybe even something like "Deep Blue" dedicated to plasma phyics.

Correlate and coordinate multiple sensor and communication channels.

Not something the Ship's Computer is expected to do: should be part of the sensor package/bridge. Depending on how closely you want these correlated it could take anything from the size of a pack of ciggaretts to a .5mx.5mx1.5m rack.

Monitor people passively through biometric observation to evaluate threat potential in a wide range of situations.

Depends more on your internal sensor package, Easily handled by a desktop computer.

Monitor and balance the life support of several people.

Not a CT job. The LSS doesn't go TU if the computer is destroyed. But local processors equivalent to a GHz laptop will be needed.

Monitor and balance the status of several critical care level patients.

Not a function of a CT ships computer. Requires a local processor in the hospital bed.

Manage tactical combat across light seconds.

This will take some power. Say a shoebox sized processor.

And all of it 24/7 for a century or more with minimal maintenance, often without any maintenance except annually, with never a failure. No blue screen of death, nor even a file not found. No corruption of files, no loss of data, and any file in the computer you want is retrieved instantly.

So they are all running Unix. Otherwise you can build in redundany pretty easily.

Please include the cooling needed and reasonable access for the rarely needed mainenance.

Using discrete processors allows them to be widely seperated for cooling and maintenance.

Also make it immune to changes in atmospheric pressure from vacuum to at least a full atmosphere, and not subject to humidity changes.

And it has to be built hardened and redundant enough to take damage from a near missile or laser blast and still perform without interruption.

Trivial with integrated circuits and solid-state construction

Do that and you have a CT model/1 except for the real-space and jump-space avionics and navigation requirements. The first is negligible, the second we don't have a real clue. That's why I chose to ignore them in the requirements.

Actually, you are demanding a great deal more than CT requires from a model 1. Remember a model 1 cannot do many of the things you asked about at all and the rest it cannot do simultaneously. All it can do at one time is normal space navigation, which you admit is trivial. That program has to be unloaded before the computer can be used for anything else.

Let's put your budget in the neighbourhood of US$ million.
More like $60K. Might be a lot less

 

[Jame]

Originally posted by Sigg Oddra:
The reason I listed the computer tonnage as negligable in the earlier post is that I think it can be safely "hidden" within the bridge tonnage.

The computer's complexity and ability to run more demanding software should increase as TL increases, but, as Jamie has put it in the past, a TL9 model 1 computer should be a lot smaller than the TL5 model 1.

I wonder if the TL column on the computer table needs a revision?

If the model 1 became TL 7, with bis and fib models just costing more money rather than requiring a higher TL as well, then the TL progression remains the same once you get above the model 2.

Well, that's getting a bit more technical than I had planned, but basically, yes. (Keep in mind that I hadn't thought much besides my initial statement.)

The other reason that I wanted to significantly reduce size is that a Tl 9 model/1 will have an extremely better display model than a Tl 5 one.

An example: the computer I'm using, and expect that most if not all of you are using, is (even though it could be a model/1) small enough to fit on top of my desk, has a monitor that can display alphanumeric text and translate the software necessary to view CotI. The controls are a keyboard, two buttons (monitor- and hard disk-power) and the knob for the speakers.

In Charlestown (MA/USA), there is a WW2-era destroyer named the Cassin Young, and its computer takes up a space roughly the size of my dining room (no, I don't know the dimensions, but I'm sure they look the same). This computer is good only for plotting trajectories and fire-control data for the main guns. Its controls consist entirely of knobs and switches, and it doesn't have a single monitor or bit of software, nor keyboard.

Theoretically, both are model/1s, but mine is a quantum leap beyond the destroyers.

The kind of computer I'd expect to run things for a starship would be a server roughly the size of the desk upon which my PC rests, and that's maybe three feet squared (bear with me; I have no intention of measuring). And the purpose of this server would, I imagine, be to centralize all the various, specialized desktop-type computers that take care of all the various functions, i.e. sensors and life support, which would have some processors of their own so as to forgo reliance on the central processor. It won't weigh one ton - it'll weigh less, and the rest of that ton could be sensors, wiring and access space.

Maybe I'm running a bit long, and I'm sure I forgot something, but this is pretty much how I feel.

Sigg, unless you're using "Jamie" as some sort of verbal gesture (go ahead, but I'd like to know now), there's no "i" in my name.

 

[Jame]

Originally posted by Sigg Oddra:
The reason I listed the computer tonnage as negligable in the earlier post is that I think it can be safely "hidden" within the bridge tonnage.

The computer's complexity and ability to run more demanding software should increase as TL increases, but, as Jamie has put it in the past, a TL9 model 1 computer should be a lot smaller than the TL5 model 1.

I wonder if the TL column on the computer table needs a revision?

If the model 1 became TL 7, with bis and fib models just costing more money rather than requiring a higher TL as well, then the TL progression remains the same once you get above the model 2.

Well, that's getting a bit more technical than I had planned, but basically, yes. (Keep in mind that I hadn't thought much besides my initial statement.)

The other reason that I wanted to significantly reduce size is that a Tl 9 model/1 will have an extremely better display model than a Tl 5 one.

An example: the computer I'm using, and expect that most if not all of you are using, is (even though it could be a model/1) small enough to fit on top of my desk, has a monitor that can display alphanumeric text and translate the software necessary to view CotI. The controls are a keyboard, two buttons (monitor- and hard disk-power) and the knob for the speakers.

In Charlestown (MA/USA), there is a WW2-era destroyer named the Cassin Young, and its computer takes up a space roughly the size of my dining room (no, I don't know the dimensions, but I'm sure they look the same). This computer is good only for plotting trajectories and fire-control data for the main guns. Its controls consist entirely of knobs and switches, and it doesn't have a single monitor or bit of software, nor keyboard.

Theoretically, both are model/1s, but mine is a quantum leap beyond the destroyers.

The kind of computer I'd expect to run things for a starship would be a server roughly the size of the desk upon which my PC rests, and that's maybe three feet squared (bear with me; I have no intention of measuring). And the purpose of this server would, I imagine, be to centralize all the various, specialized desktop-type computers that take care of all the various functions, i.e. sensors and life support, which would have some processors of their own so as to forgo reliance on the central processor. It won't weigh one ton - it'll weigh less, and the rest of that ton could be sensors, wiring and access space.

Maybe I'm running a bit long, and I'm sure I forgot something, but this is pretty much how I feel.

Sigg, unless you're using "Jamie" as some sort of verbal gesture (go ahead, but I'd like to know now), there's no "i" in my name.

 

[mike wightman]

Sigg, unless you're using "Jamie" as some sort of verbal gesture (go ahead, but I'd like to know now), there's no "i" in my name.

Please accept my humble apologies sir.

I have a nephew named Jamie, so it is force of habit.

I'll edit accordingly and try to get it right in future.

Why isn't all of this computer stuff going into the CT+ computer thread?

 

[mike wightman]

Sigg, unless you're using "Jamie" as some sort of verbal gesture (go ahead, but I'd like to know now), there's no "i" in my name.

Please accept my humble apologies sir.

I have a nephew named Jamie, so it is force of habit.

I'll edit accordingly and try to get it right in future.

Why isn't all of this computer stuff going into the CT+ computer thread?

 

[Jame]

Oh, a nephew. That's alright; now that I know, I can forgive . As for the computer stuff, well, I'll just have to put it in there, now won't I?

 

[Jame]

Oh, a nephew. That's alright; now that I know, I can forgive . As for the computer stuff, well, I'll just have to put it in there, now won't I?