Starship Computers in v. 3

[Gruffty the Hiver]

...And, as if by magic, the thread appears...

My idea for computers in this version of Traveller:

Each major system of the ship has its own computer (as has already been suggested in the other thread). Each computer (that is attached to each component) has a "rating" number - say, perhaps, a J-6 drive has a Jump Computer-6 or a Maneuver-2 drive has a Maneuver Computer-2, and so on. Add up all the "ratings" within the ship, look up the result on a table in the book and, voila! you have your overall starhsip computer model number.

There's even potential to make this backwardly compatible with CT/HG, so that a specific overall computer "rating" relates to a specific computer in CT/HG.

 

[Lorcan Nagle]

It depends on how much complexity you want to bring in - having seperate computers gives rise to concepts like networking the system, and a sufficiently skilled programmer could take control of navigation via the jump computer and so on. Interesting story ideas there for sure.

From a practical standpoint, the ship would likely have a governing computer for each major system, and central system to co-ordinate them (a server-type setup). To keep compatibility with previous iterations of Traveller you just have to match mass and cost while giving them whatever capacities you want to flesh them out more.

 

[Gruffty the Hiver]

Another way of working computers would be to assign a % value for each component and the hull and add them up (simple and quick, no extra tables in the book needed). So a 100 ton ship with Jump Drive-A (10 tons), Maneuver Drive-A (2 tons) and Power Plant-A (4 tons) would need a computer capable of handling 100+10+2+4 = 116 tons, which would occupy a+b+c+d% of the hull and cost w+x+y+z MCr per ton of computer.

Thought: divide the hull+Jump+Maneuver+power plant tonnages by 100 to get the computer size. In the example above that would give a computer that occupies 1.16 tons (this could be rounded up or down to save on maths during ship construction).

 

[Gruffty the Hiver]

It depends on how much complexity you want to bring in

Yeah, keeping it simple and not having it end up as another version of Fire, Fusion & Steel is a problem!

having seperate computers gives rise to concepts like networking the system, and a sufficiently skilled programmer could take control of navigation via the jump computer and so on. Interesting story ideas there for sure.

Indeed!

From a practical standpoint, the ship would likely have a governing computer for each major system

I agree.

and central system to co-ordinate them (a server-type setup). To keep compatibility with previous iterations of Traveller you just have to match mass and cost while giving them whatever capacities you want to flesh them out more.

That's similiar to T20 in approach but worth re-visiting.

 

[Gruffty the Hiver]

OK, from High Guard (reproduced here in simplified form and soley for discussion purposes, (c) GDW/FFE):

Computer  USP
Model    Code  MCr  Tons  Capacity       Hull  TL  EP
1           1    2     1      2/ 4        600   5   0 
1 fib       A    3     1      2/ 4        600   5   0
1 bis       R    4     1      4/10        600   6   0
2           2    9     2      3/ 6      1,000   7   0
2 fib       B   14     4      3/ 6      1,000   7   0
2 bis       S   18     2      6/ 0      1,000   8   0
3           3   18     3      5/ 9      4,000   9   1
3 fib       C   27     6      5/ 9      4,000   9   1
4           4   30     4      8/15     10,000  10   2
4 fib       D   45     8      8/15     10,000  10   2
5           5   45     5     12/25     50,000  11   3
5 fib       E   68    10     12/25     50,000  11   3
6           6   55     7     15/35    100,000  12   5
6 fib       F   83    14     15/35    100,000  12   5
7           7   80     9     20/50  1,000,000  13   7
7 fib       G  100    18     20/50  1,000,000  13   7
8           8  110    11     30/70          -  14   9
8 fib       H  140    22     30/70          -  14   9
9           9  140    13     40/90          -  15  12
9 fib       J  200    26     40/90          -  15  12

Note:
Capacity indicates CPU/storage;
Hull is the hull requiring this computer as a minimum;
TL is Tech Level;
EP is the computer energy point requirement.

The v. 3 playtest document states that EMP hardening increases the computer's cost by 50%, but HG doesn't do this with the fib(re optic) computers - it doubles the size of the computer in tons, and adds a bit to the cost.

Computer
Model     TL  Rating  MCr
1          7       5    0.03
2          9      10    0.16
3         11      15    2
4         12      20    5
5         13      25   10
6         14      30   20
7         15      35   30

Options:

Jump Control Specialisation (BIS models): A computer's rating can be increased by 5 for the purposes of running Jump
Control programs only. This increases the computer's cost by 50%.

Hardened Systems: A computer and its connections can be hardened against attack, ensuring the system is protected
against electromagnetic pulse weapons. A hardened system is immune to EMP, but costs 50% more.

If you remove all bis and fib models from the HG table, and the USP Code, Capacity and EP columns then adjust the TLs for the lower model computers, this is what you get:

Computer
Model    MCr  Tons    Rating       Hull  TL
1          2     1         ?        600   7
2          9     2         ?      1,000   8
3         18     3         ?      4,000   9
4         30     4         ?     10,000  10
5         45     5         ?     50,000  11
6         55     7         ?    100,000  12
7         80     9         ?  1,000,000  13
8        110    11         ?          -  14
9        140    13         ?          -  15

 

[Gruffty the Hiver]

HG adds approximately 50% of the basic computer cost and doubles the tonnage of the computer for fib models.

 

[Allensh]

I like the fact that don't even mention computer sizes. Avoids the entire issue and subsumes that into the overall size of the bridge, even though they tell you the computers are located near the bridge but not on the bridge.

Since they therefore are not using computer sizes, it makes sense to modify the cost instead. 50% for fib might be a bit high though, maybe 20% would be better.

Allen

 

[Gruffty the Hiver]

I think I'm having one of those "I put my brain down somewhere and can't find it" days today.....

Yes, ignore everything I said above :roll:

 

[Golan2072]

I usually treat the Traveller "computer" as an abstraction of the ship's electronic systems, from communication relays to sensors to network hardware to actual computers. This way I avoid the "networking vs. mainframe" issue and the "where is the sensors/comms tonnage" problem.

So I suggest noting something similar in the rules to avoid the famous issues of Traveller computing - note that this is an abstraction, and if you want it to be one big mainframe let it be, if you want it to be an integrated network let it be, the rules could be interpreted into both.

 

[pasuuli]

I like the starship-computer-network idea. Especially if I don't have to worry about buying the proper software packages to get my jump drive to work.

 

[GamerDude]

I like the idea of the networked distributed systems... but also understand that if you are buying only one (in terms of the system rules) computer then to me it's a central system with smart terminals handling things like interfacing and such.

but, either way how ever it works for you then it works for you.

 

[pasuuli]

I like the idea of the networked distributed systems... but also understand that if you are buying only one (in terms of the system rules) computer then to me it's a central system with smart terminals handling things like interfacing and such.

but, either way how ever it works for you then it works for you.

Well regardless of what it is, I would not want to have to buy a computer for every component!

 

[silburnl]

I like the fact that don't even mention computer sizes. Avoids the entire issue and subsumes that into the overall size of the bridge, even though they tell you the computers are located near the bridge but not on the bridge.

I noticed this when I was going back through this section last night and wondered if it was bug, but on reflection I agree with you that it's a nice feature and usefully sidesteps the 'computer weighs what!' bugbear of CT.

This new scheme ties computer capability to a combination of TL and price, which seems a bit bizarre - I'd have thought that you would have one axis of difference or the other, but not both (unless you introduced a full matrix of combinations, which seems more like the approach to be adopted for a naval architect's supplement). My first reaction to the this table was 'OK so if fitting a type 1 computer from That 70s Planet costs 0.3MCr, how much would it cost to get one of the same spec from Rhylanor?'

Or is the assumption that all ships are built at TL15 shipyards and the TL column is there simply to indicate which rows to eliminate if you are playing in a tech-limited setting?

Regards
Luke

 

[qstor]

I still don't understand why a scout has such a poor computer system. The sensor packages on a Model 1 are barely enough to survey a planet let alone a whole system.

Skimming the playtest doc it seems Mongoose Traveller is little more than a CT rehash.

Mike

 

[GamerDude]

I like the idea of the networked distributed systems... but also understand that if you are buying only one (in terms of the system rules) computer then to me it's a central system with smart terminals handling things like interfacing and such.

but, either way how ever it works for you then it works for you.

Well regardless of what it is, I would not want to have to buy a computer for every component!

You only buy "one" computer, but you have to purchase the programs to run what's on your ship.

And, for the "why isn't this a 20th century car w/firmware built in", think more of mid-frames and mainframes, which is what you are getting for your ship...

Ya don't get the damm software with it except for the operating system. Everything else costs and the same system can be used in many environments from businesses, production plants, starships, scientific research/data modeling, etc.

 

[pasuuli]

I like the idea of the networked distributed systems... but also understand that if you are buying only one (in terms of the system rules) computer then to me it's a central system with smart terminals handling things like interfacing and such.

but, either way how ever it works for you then it works for you.

Well regardless of what it is, I would not want to have to buy a computer for every component!

You only buy "one" computer, but you have to purchase the programs to run what's on your ship.

And, for the "why isn't this a 20th century car w/firmware built in", think more of mid-frames and mainframes, which is what you are getting for your ship...

Ya don't get the damm software with it except for the operating system. Everything else costs and the same system can be used in many environments from businesses, production plants, starships, scientific research/data modeling, etc.

So, you mean computers are computers, this one just happens to be for a starship. That's quite reasonable.

 

[Takei]

You only buy "one" computer, but you have to purchase the programs to run what's on your ship.

And, for the "why isn't this a 20th century car w/firmware built in", think more of mid-frames and mainframes, which is what you are getting for your ship...

Ya don't get the damm software with it except for the operating system. Everything else costs and the same system can be used in many environments from businesses, production plants, starships, scientific research/data modeling, etc.

Being the main proponent of the distributed computing with firmware built in argument I will say that it is very easy to just ignore computers and software. For those that like the extra complexity use the computers and software presented as is. I think I'd like to see the computer-less approach mentioned in a sidebar.

From a 'realism' point of view it would still be interesting to talk to people from the aerospace and shipbuilding industries to see how control of the vehicles is handled.

EDIT: Added the following;

One argument ignored by the software supporters though is the question of character skill. If ship's weapons are controlled by FireControl/x where's the skill in moving your curser over the target to designate it? You've effectively negated any gunnery skill over the level 1 needed to know the basics of operating the system.
The same applies to the Evade program; does it negate Pilot skill? Can you use Pilot skill whilst running Evade/x, or does the computer take over when it detects an attack? If it doesn't then surely you can't get the benefit of the program?

 

[Lorcan Nagle]

Expert systems could be modelled as being very predictable, maybe having a penalty in initiaitve systems and a human pilot could make a tactics roll to realise they're fighting an expert system and switch tactics to fool it.

 

[atpollard]

One argument ignored by the software supporters though is the question of character skill. If ship's weapons are controlled by FireControl/x where's the skill in moving your curser over the target to designate it? You've effectively negated any gunnery skill over the level 1 needed to know the basics of operating the system.
The same applies to the Evade program; does it negate Pilot skill? Can you use Pilot skill whilst running Evade/x, or does the computer take over when it detects an attack? If it doesn't then surely you can't get the benefit of the program?

Don't think of it as a Video Game or Engineering software with absolute results. Think of it like a weather model for forecasting Hurricane tracks . There are too many variables that cannot be known with certainty, so the skill of the 'Operator' in adjusting the estimated variables determines the final accuracy of the prediction.

 

[Takei]

Don't think of it as a Video Game or Engineering software with absolute results. Think of it like a weather model for forecasting Hurricane tracks . There are too many variables that cannot be known with certainty, so the skill of the 'Operator' in adjusting the estimated variables determines the final accuracy of the prediction.

My engineering background was coming to the fore
The hurricane predicting sounds like a better way to looking at it.