As I've been slowly shifting my setting to be more 'cyber' and 'ghost in the shell'-like, I've had to focus attention on MT's computer rules, and I don't think they work very well. I do like the entire 'control panel' ideas for determining what's needed and its relationship to manning requirements, but the table and use in combat is not very good.
The computer table seems to link tech level with model number in such a way as to prevent things like a model 0 computer built at tech 13, and what happens when you want a model 9 computer but don't need to connect 100,000,000 panels to it; can't it be made smaller if you only need to connect to 100 panels? The table is not very flexible.
All flying vehicles, such as your typical air-raft convertible generally have 2 model 0 computers in them making them stupidly expensive...and just to handle a handful of controls; why can't they be made smaller? ( the RAW don't require 2 computers, although they suggest is a good idea for safety. In fact, unless you are using linked panels, computers aren't even necessary.)
The way computers are used in the starship combat section is ridiculous. Back in CT and HG, computers were an abstract that encompassed computers, sensors, communications, avionics, ECM, etc. where a computer DM covered the effects of all those things, but in MT, sensors and commo and ECM and avionics, etc. were factored out, yet the combat system , based on HG, still uses the computer DM in the same way making the computer more important than sensors or ECM. This is why a fighter with limited volume cannot hit a battleship with volume to spare no matter how well it rolled its sensor lock task, and that the battle ship cannot miss. I suspect that a rational update in the computer rules would change that. I intend to rewrite the space combat rules to suit me at a later date anyways. ( computer number as a dm to hit?...horseshit...how about the 'effect' from a lock-on task instead )
Computers are rated primarily by the number of CP's it can handle and the CP multiplier. I will assume that the computer can serve the number of CP's connected without lag, real-time, with dedicated cpu's represented by the CP multiplier number. This would give a model 0 tech 8 computer ( cp max=500, cp mult=5 ) a total of 2500 cpu's. By dividing stats like vol, power, cost and mass by this number, we can see a single cpu's stats. For tech-8, this makes a single cpu have a volume of .2 liters, the same as a single serial processor from CT book8 Robots.
So how about designing computers in general like CT Book 8 did them for robot brains? That'd make a tech-8 model-0 computer with a single connection have 5 cpu's for a CPmultiplier of 5 and which might represent the hardware needed to run low data and basic command with 1 cpu unit left over for other programming. IIRC, this somewhat mirrors the old CT ship computer rules. Add more memory and off-line storage as desired to keep a library of useful programs and data handy. Give each system its own dedicated computer to connect to a central main computer. Damage to computers reduces the number of CP's it can handle which takes ship systems off-line until a back-up connects and those CP's can be serviced again.
A computer's model number would be based on the number of dedicated cpu's for each process. Or perhaps using Book 8's method for determining a computer's 'intelligence' and 'education' as stats to be used as ordinary stat dm's as per the standard task rules ( in my house rules, related to stat/3 ).
A large system with many connection would have a limited number of cpu's per connection; a mainframe. A large system with few connection could have many, many cpu's dedicated to a few connections; a super-computer.
Things really start to take off when parallel processing and neural chips as envisioned by Traveller come into use where a single chip would have the computing power of several serial cpu's.
If we use the same idea with higher tech computers, cpu volume drops to mere millilitres which coincides pretty well with implanted computers people and cyber brains. Add as many as you need to have the desired computing power and programming space.
For AI, just use Robots as a guide, as well as for specific hardware additions, such as master/slave units, for tele-presence or a method of controlling exo-skeletons, such as batteldress, and radio connections to other computers without hardlines.
A rule method for hacking needs to be made, probably using programs as weapons and armor and various skills ( stealth, recon, lockpicking, forgery, et al.).
These would be rated according to the skill of the programmer with stat dm's of the computer the programs are being run on. Hacking from behind a firewall increases difficulty for each firewall operating through. Hacking directly , 'in-the-open' leads to damage taken by hacker's system directly as opposed to using firewalls as armor.
Direct neural interfaces allow for the user's brain to be used, aka wetware, such that the user's stats ans skills can be used in addition to those of the programming. Damage taken is to the user's intelligence stat with the danger of being made brain-dead.
The computer table seems to link tech level with model number in such a way as to prevent things like a model 0 computer built at tech 13, and what happens when you want a model 9 computer but don't need to connect 100,000,000 panels to it; can't it be made smaller if you only need to connect to 100 panels? The table is not very flexible.
All flying vehicles, such as your typical air-raft convertible generally have 2 model 0 computers in them making them stupidly expensive...and just to handle a handful of controls; why can't they be made smaller? ( the RAW don't require 2 computers, although they suggest is a good idea for safety. In fact, unless you are using linked panels, computers aren't even necessary.)
The way computers are used in the starship combat section is ridiculous. Back in CT and HG, computers were an abstract that encompassed computers, sensors, communications, avionics, ECM, etc. where a computer DM covered the effects of all those things, but in MT, sensors and commo and ECM and avionics, etc. were factored out, yet the combat system , based on HG, still uses the computer DM in the same way making the computer more important than sensors or ECM. This is why a fighter with limited volume cannot hit a battleship with volume to spare no matter how well it rolled its sensor lock task, and that the battle ship cannot miss. I suspect that a rational update in the computer rules would change that. I intend to rewrite the space combat rules to suit me at a later date anyways. ( computer number as a dm to hit?...horseshit...how about the 'effect' from a lock-on task instead )
Computers are rated primarily by the number of CP's it can handle and the CP multiplier. I will assume that the computer can serve the number of CP's connected without lag, real-time, with dedicated cpu's represented by the CP multiplier number. This would give a model 0 tech 8 computer ( cp max=500, cp mult=5 ) a total of 2500 cpu's. By dividing stats like vol, power, cost and mass by this number, we can see a single cpu's stats. For tech-8, this makes a single cpu have a volume of .2 liters, the same as a single serial processor from CT book8 Robots.
So how about designing computers in general like CT Book 8 did them for robot brains? That'd make a tech-8 model-0 computer with a single connection have 5 cpu's for a CPmultiplier of 5 and which might represent the hardware needed to run low data and basic command with 1 cpu unit left over for other programming. IIRC, this somewhat mirrors the old CT ship computer rules. Add more memory and off-line storage as desired to keep a library of useful programs and data handy. Give each system its own dedicated computer to connect to a central main computer. Damage to computers reduces the number of CP's it can handle which takes ship systems off-line until a back-up connects and those CP's can be serviced again.
A computer's model number would be based on the number of dedicated cpu's for each process. Or perhaps using Book 8's method for determining a computer's 'intelligence' and 'education' as stats to be used as ordinary stat dm's as per the standard task rules ( in my house rules, related to stat/3 ).
A large system with many connection would have a limited number of cpu's per connection; a mainframe. A large system with few connection could have many, many cpu's dedicated to a few connections; a super-computer.
Things really start to take off when parallel processing and neural chips as envisioned by Traveller come into use where a single chip would have the computing power of several serial cpu's.
If we use the same idea with higher tech computers, cpu volume drops to mere millilitres which coincides pretty well with implanted computers people and cyber brains. Add as many as you need to have the desired computing power and programming space.
For AI, just use Robots as a guide, as well as for specific hardware additions, such as master/slave units, for tele-presence or a method of controlling exo-skeletons, such as batteldress, and radio connections to other computers without hardlines.
A rule method for hacking needs to be made, probably using programs as weapons and armor and various skills ( stealth, recon, lockpicking, forgery, et al.).
These would be rated according to the skill of the programmer with stat dm's of the computer the programs are being run on. Hacking from behind a firewall increases difficulty for each firewall operating through. Hacking directly , 'in-the-open' leads to damage taken by hacker's system directly as opposed to using firewalls as armor.
Direct neural interfaces allow for the user's brain to be used, aka wetware, such that the user's stats ans skills can be used in addition to those of the programming. Damage taken is to the user's intelligence stat with the danger of being made brain-dead.
