Computers/Fire control for medium range ships (600-2000 dtons)

[whartung]

Towards that end to execute successful evades I think the ships are maneuvering with heavy use of roll pitch and yaw plus some moves 'up' or 'down' relative to the plane between firing ships and the target.

…

A miss also doesn't have to be a clean miss, having the ship be on an unexpected roll spreading the hit across many more square meters of hull or hitting the edge of the ship such that sloped armor effects start occurring can avoid damage.

Lasers, PA, and Meson Guns are "burst" weapons. Witness the X-Ray Detonation laser. Lasers do not kill things by sitting on them in one place. They overwhelm the target with a burst of energy, or they bounce off. It's hard enough to deliver enough energy per cm^2 in the first place to do any damage at range, much less rely up sitting on the exact same cm^2 and waiting for it to melt.

So, rolling doesn't do anything (save perhaps randomly change a hit location). If rolling could make a difference, so does relative velocity of the attacker, for the same reasons. I'd hate to be in a ship that's rolling at a rate of 1000's m/s, gravitic compensators or no.

Every energy weapon shot is as a ballistic computation like most any slug thrower shot. Even gravity effects the trajectory. The gunnery computers are not making "snap shots", they're making well planned ballistic projections based on a long history of analyzing the ships vectors an behaviors. The computers are making these decisions at a sub-millisecond scale (even if it takes 100's of ms to execute the decision). So, while things look like they're happening fast to us meatspace carbon based lifeforms, the computers see everything in "slow motion".

Consider: https://www.youtube.com/watch?v=ZVNnoOcohaU

Simply, all of the reactions that your ship must do in order to defeat the computers prediction system must happen within the reaction time of the computer system. Your thrust changes have to happen faster than the computer. Your thruster directional changes have to happen faster (and remember, your ship has momentum to be overcome for your thrusters to start it's maneuver, as well as overcome to stop that maneuver -- and those are all predictable).

The best you can do is random thrust inputs, random acceleration bursts, and perhaps random gimbal motions (depending on the response time of the gimbals). On larger ships, I don't think this is enough to matter. Smaller ships may get some benefit. But even still, it's been argued that an attack is a bunch of bursts over a period of time, thus filling the projection window with fire.

The firing solution is watching your every move, and predicting all of your possible outcomes 2-4s out in advance. It's watching you like a hawk. Then it fills that projected area with energy in hopes that you'll intersect with some of it.

And if you ever stop, if your random inputs stop for any reason (such as a dramatic, controlled vector change), the projection window gets very small, very fast.

All of the rest is noise in the system. The latency within the firing ship does not weaken the solution directly, it simply allows the input of noise from other sources (such as random thrust inputs by the target) to possibly expand the projected window. With no noise, it has no impact whatsoever.

Also, note, that random inputs by the evading ship don't necessarily undermine it's firing solutions back at you. It KNOWS what it's going to do. While each thrust input may be random to the observer, the control computers already know when and how much each burst will be, and can be compensated in the solution.

 

[kilemall]

Lasers, PA, and Meson Guns are "burst" weapons. Witness the X-Ray Detonation laser. Lasers do not kill things by sitting on them in one place. They overwhelm the target with a burst of energy, or they bounce off. It's hard enough to deliver enough energy per cm^2 in the first place to do any damage at range, much less rely up sitting on the exact same cm^2 and waiting for it to melt.

So, rolling doesn't do anything (save perhaps randomly change a hit location). If rolling could make a difference, so does relative velocity of the attacker, for the same reasons. I'd hate to be in a ship that's rolling at a rate of 1000's m/s, gravitic compensators or no.

We'll have to agree to disagree on this one, even assuming all energy force arrives at the same few milliseconds a roll might be the difference between a penetration and not.

Spinal weapon hit, eh, probably not, but I have rules to deal with that situation.

Relative velocity for energy weapons? Hardly.

Now then, if you will recall the missile supplement DOES have a relative speed kinetic impact rule, so in that case relative velocity which would in include large measure imparted vee from the launching vessel DOES count, and if that rule is honored certainly other impactors such as from a mass driver should gain 'bonuses' to penetration and damage from initial launch vee, if it is high enough.

Every energy weapon shot is as a ballistic computation like most any slug thrower shot. Even gravity effects the trajectory. The gunnery computers are not making "snap shots", they're making well planned ballistic projections based on a long history of analyzing the ships vectors an behaviors. The computers are making these decisions at a sub-millisecond scale (even if it takes 100's of ms to execute the decision). So, while things look like they're happening fast to us meatspace carbon based lifeforms, the computers see everything in "slow motion".

Consider: https://www.youtube.com/watch?v=ZVNnoOcohaU

Simply, all of the reactions that your ship must do in order to defeat the computers prediction system must happen within the reaction time of the computer system. Your thrust changes have to happen faster than the computer. Your thruster directional changes have to happen faster (and remember, your ship has momentum to be overcome for your thrusters to start it's maneuver, as well as overcome to stop that maneuver -- and those are all predictable).

The best you can do is random thrust inputs, random acceleration bursts, and perhaps random gimbal motions (depending on the response time of the gimbals). On larger ships, I don't think this is enough to matter. Smaller ships may get some benefit. But even still, it's been argued that an attack is a bunch of bursts over a period of time, thus filling the projection window with fire.

The firing solution is watching your every move, and predicting all of your possible outcomes 2-4s out in advance. It's watching you like a hawk. Then it fills that projected area with energy in hopes that you'll intersect with some of it.

And if you ever stop, if your random inputs stop for any reason (such as a dramatic, controlled vector change), the projection window gets very small, very fast.

Precisely my understanding of the situation, and what I think evade is exactly doing.

My CT/HG system will incorporate firing patterns and effects of tradeoffs between generating a hit and damage.

All of the rest is noise in the system. The latency within the firing ship does not weaken the solution directly, it simply allows the input of noise from other sources (such as random thrust inputs by the target) to possibly expand the projected window. With no noise, it has no impact whatsoever.

Agree to disagree then.

Also, note, that random inputs by the evading ship don't necessarily undermine it's firing solutions back at you. It KNOWS what it's going to do. While each thrust input may be random to the observer, the control computers already know when and how much each burst will be, and can be compensated in the solution.

Of course, which is why it is utterly imperative that you have one data input/control point knowing precisely the state of your ship AND the latest on the target ship.

 

[kilemall]

Alright then, as to Mr. Wightman's ruling, I am going to load a Model/5 computer to the gills with every tasty program one would want, short again of every jump program (which is wasted space 98% of the time).

2 Predict-5 *
1 Gunner Interact
1 Target
1 Select-3
4 Multi-Target 4
1 Launch
4 Double Fire
3 Maneuver/Evade 6
1 Return Fire
2 Anti Missile
3 ECM
6 Jump-6
1 Navigation
1 Generate
1 Anti Hijack
1 Library

33 spaces, a Mod/5 has 12/25 for 37 slots, and with a big Jump 6 program in there- like I said, you literally don't need a bigger machine other then the ship build rules.

Whenever assessing a creation like a game rule I'm looking for consistency of intended effect- even if it is not achieved or is a questionable implementation, intent is usually there.

I just don't believe that the game design was intended to taper off the utility or need for bigger computers for bigger ships, but that consistency of the shortage of computing resources for every desirable thing and hard choices is still there at 2000 tons like it was at 200.

And GI is the variable that puts the demand on the machine.

Course, here is the really fun part to consider- this same Mod/5 or it's larger brethren don't really need any more programs then the above.

So those are all the programs required to run a 300,000 ton superdreadnaught?

 

[McPerth]

Alright then, as to Mr. Wightman's ruling, I am going to load a Model/5 computer to the gills with every tasty program one would want, short again of every jump program (which is wasted space 98% of the time).

2 Predict-5 *
1 Gunner Interact
1 Target
1 Select-3
4 Multi-Target 4
1 Launch
4 Double Fire
3 Maneuver/Evade 6
1 Return Fire
2 Anti Missile
3 ECM
6 Jump-6
1 Navigation
1 Generate
1 Anti Hijack
1 Library

<snip>

While this small detail does not take any rightnes of your post at large, see that to be able to perform jump 6 (as hinted in your program selection) you need a computer 6, as jump may not be higher than your computer number.

That means that probably computer 6 will be relegated to ships with J6 capability, that don't use to be superdreadnaughts, but more likely courriers.

 

[mike wightman]

Alright then, as to Mr. Wightman's ruling, I am going to load a Model/5 computer to the gills with every tasty program one would want, short again of every jump program (which is wasted space 98% of the time).

2 Predict-5 *
1 Gunner Interact
1 Target
1 Select-3
4 Multi-Target 4
1 Launch
4 Double Fire
3 Maneuver/Evade 6
1 Return Fire
2 Anti Missile
3 ECM
6 Jump-6
1 Navigation
1 Generate
1 Anti Hijack
1 Library

33 spaces, a Mod/5 has 12/25 for 37 slots, and with a big Jump 6 program in there- like I said, you literally don't need a bigger machine other then the ship build rules.

Whenever assessing a creation like a game rule I'm looking for consistency of intended effect- even if it is not achieved or is a questionable implementation, intent is usually there.

I just don't believe that the game design was intended to taper off the utility or need for bigger computers for bigger ships, but that consistency of the shortage of computing resources for every desirable thing and hard choices is still there at 2000 tons like it was at 200.

And GI is the variable that puts the demand on the machine.

Course, here is the really fun part to consider- this same Mod/5 or it's larger brethren don't really need any more programs then the above.

So those are all the programs required to run a 300,000 ton superdreadnaught?

They have to be active to be used in your turn, so you have 12 slots available during your turn. Yes, you can switch between active and storage during the computer phase, but because you could be the target of laser return fire and missile impacts you need to be running your maneuver/evade and other defensive programs as well as your offensive programs.

So for a warship:
predict 5 2
gunner I 1
target----1
launch---1
m/e6-----3
return fire 1
antimissile 2
ECM 3

I make that 14 which is beyond a model 5, a model 6 is required.

Offensively you may want
select 3 1
multi----1,2 or 4
double fire 4

That's at least an extra 6 which requires a model 7 and yet still can't run multi 2 or 3 without dropping double fire - hence a trade off.

 

[jawillroy]

For what it's worth, ever since I first got the LBBs as a cheese-eating-teenager in the eighties, I'd assumed that only one GI program was necessary. I'm with Wightman on this one.

 

[kilemall]

They have to be active to be used in your turn, so you have 12 slots available during your turn. Yes, you can switch between active and storage during the computer phase, but because you could be the target of laser return fire and missile impacts you need to be running your maneuver/evade and other defensive programs as well as your offensive programs.

So for a warship:
predict 5 2
gunner I 1
target----1
launch---1
m/e6-----3
return fire 1
antimissile 2
ECM 3

I make that 14 which is beyond a model 5, a model 6 is required.

Offensively you may want
select 3 1
multi----1,2 or 4
double fire 4

That's at least an extra 6 which requires a model 7 and yet still can't run multi 2 or 3 without dropping double fire - hence a trade off.

Incorrect ruling then IMO, sirrah.

LBB2, Page 38 paragraph 3 (giving mod/1 example)

Of these six, only two (the capacity limit of the CPU) can function at any one time (in one phase). In the laser return fire phase, both target and return fire programs would be required, and only those programs could be used with this capacity CPU to effect laser return fire. During a laser fire phase, as before, only two programs could be used; target is required, but the player could select between predict-1 or gunner interact for the program to be processed, depending on which would provide the greater benefit.

Next paragraph addresses the computer reprogramming phase, with the example of dropping programs to allow a load of jump-related programs.

So given the phase specific wording, no you DON'T have to be running the maneuver/evade or ECM/antimissile, they would be running during the enemy laser fire/missile impact phases respectively.

Again, the rule makes it clear you are not running all the programs all phases, just specifically during the phase they come into play.

IMTU I actually WILL likely use capacity your way for the whole turn especially for evade and return fire since I am looking to do hardmode ship tactics fighting your ship, it's just been clear to me all along how those LBB2 computers worked capacity by phasing wise (otherwise those trader vessels are practically defenseless).

But also, I have an IMTU out for players with the cheaper computers, the darn LBB2/HG computers can practically double the ship cost (but also leverage the power of the ship, so you have to be careful with keeping price/performance/reliability in sync with each other).

It's just relevant when trying to figure out intent from vague lines like the GI usage rule.

 

[mike wightman]

It specifically states that you do to get the DMs in the required phase...
Intruder movement phase
Intruder laser fire phase - this is where you need your full defensive loadout running - m/e (3s), rf (1s), am (2s), ECM (3s) = 9 spaces
Native(you) laser return fire - this is where you need your minimum offensive loadout - target (1s), p5 (2s), gi (1s) = 4
Total 13 - model 6 computer required

All of this needs to be running in your computer as a minimum to get the best DMs - as a warship should be aiming for.

It is now almost your turn to act offensively so you could drop the ECM and am programs since they are of no use to you during your offensive turn freeing up 5 spaces for multitarget, double fire, selective, launch.

 

[kilemall]

It specifically states that you do to get the DMs in the required phase...
Intruder movement phase
Intruder laser fire phase - this is where you need your full defensive loadout running - m/e (3s), rf (1s), am (2s), ECM (3s) = 9 spaces
Native(you) laser return fire - this is where you need your minimum offensive loadout - target (1s), p5 (2s), gi (1s) = 4
Total 13 - model 6 computer required

All of this needs to be running in your computer as a minimum to get the best DMs - as a warship should be aiming for.

It is now almost your turn to act offensively so you could drop the ECM and am programs since they are of no use to you during your offensive turn freeing up 5 spaces for multitarget, double fire, selective, launch.

Mmmm, no, if its not used in the phase it goes to storage, I don't see why this is so not understandable.

Return Fire and Anti missile just alone is not required during the intruder fire phase, they will be loaded from storage and used in the return fire phase only. M/E is not used in the return fire phase, only RF AM and Target/Multi-target.

During friendly laser fire RF, AM, ECM, M/E will all be dumped into storage in favor of whatever target programs plus predict and GI programs.

I can quote many more lines in support of this interpretation, it's crystal clear, unlike the GI requirements.

 

[mike wightman]

I am an idiot.

My understanding of the rules (up to now) is that you can only move programs between active CPU and storage during the reprogramming phase.

Saying that I have just realised (thanks for making me look into this again) I have been doing it wrong for 35 years - Mayday even gives a better example of moving programs between phases from CPU to storage.

That said you will still need at least a model 6 computer on a state of the art warship to run every offensive program you can during your laser fire phase (I am clutching at straws now )

 

[McPerth]

I am an idiot.

My understanding of the rules (up to now) is that you can only move programs between active CPU and storage during the reprogramming phase.

Saying that I have just realised (thanks for making me look into this again) I have been doing it wrong for 35 years - Mayday even gives a better example of moving programs between phases from CPU to storage.

You're not the only one, if that makes you feel better. I realized it in post 2 this same thread, when kilemall was so kind as to explain it, but I was also wrong for 35 years.

 

[kilemall]

Well if it's any consolation I was already a computer operator by then.

My 'big' machine had 384K OS, 640K assignable memory in 6 partitions (virtual machines running off one processor), and prioritization between the partitions in contro by the operator. Heh, pretty much a Model/1, come to think of it.

So I had a real live machine I could compare the rules to and see the 'sense' in them about how the game computers worked.

 

[kilemall]

While this small detail does not take any rightnes of your post at large, see that to be able to perfoerm jump 6 (as hinted in your program selection) you need a computer 6, as jump may not be higher than your computer number.

That means that probably computer 6 will be relegated to ships with J6 capagility, that don't use to be superdreadnaughts, but more likely courriers.

Point taken. So Model/6 is the absolute gotta have limit with purely a 1 GI interpretation, in a CT environment. HG bonuses and build requirements may make it worthwhile otherwise.

 

[mike wightman]

HG first edition allows you to have bis models of every computer, the USP letters are still available in HG second edition but the rule isn't mentioned.

 

[kilemall]

So then, with this new understanding, for the OP's original question re: CT builds to 5000 tons, SHOULD GI be interpreted as 'per gunner', 'per turret', 'per target', or as a serve-all utility like Predict apparently does (no per target requirement for that one)?

And what if you go CT computer style all the way to 1,000,000 ton ships? What program mix is a Model/9 running that justifies the +9 HG mod?

 

[rancke]

And what if you go CT computer style all the way to 1,000,000 ton ships? What program mix is a Model/9 running that justifies the +9 HG mod?

The computer rules for basic Traveller make for good gaming, but they're hardly realistic. And the HG computer rules have their own problems where they have been simplified for game purposes. I don't think it's feasible to mix them up and get anything usable out of it. Or to draw conclusions for one set of rules based on the other set.


Hans

 

[Meteoric Assault]

Condumdrum

As I am designing and putting out plans for ships this little dynamic of Computer size and related software is becoming apparent to my own surprise.

The Berka Heavy Corvette was crippled by a 3:1 fight with three Gazelle class escorts... I think one of the two issues i discovered was the Computer software issue that was a design shortfall and the absolute Ep drain the PA barbette was and how I lost agility. Both are being rectified.

FYI the Berka was designed to stand up to such odds with ease. As I previously stated I am sacrificing Jump4 for more room for Armor(to factor 3), a larger computer model ( Model 6), and a second power plant to boost EP (boost to cover he 6 EPs for the PA and to give an agility of 5).

HG Stats of the redesign-
LH-00229 Berka LH-8235662-340000-40103-1 MCr 723.03 800.00 tons

 

[kilemall]

The computer rules for basic Traveller make for good gaming, but they're hardly realistic. And the HG computer rules have their own problems where they have been simplified for game purposes. I don't think it's feasible to mix them up and get anything usable out of it. Or to draw conclusions for one set of rules based on the other set.


Hans

My target is not drawing conclusions, its running HG sized ships in a modified CT game, including movement and crew drama.

 

[mike wightman]

There is a rule in the computer section that referees and players are encouraged to make up their own computer programs.

One that stands out as a necessity for large ships is a battery program that links turrets to be controlled by a single gunner.

Then there are the control programs needed to use nuclear dampers and meson screens.

A series of bay weapon programs, spinal mount programs, higher ratings for predict and maneuver/evade - that sort of thing.

 

[kilemall]

There is a rule in the computer section that referees and players are encouraged to make up their own computer programs.

One that stands out as a necessity for large ships is a battery program that links turrets to be controlled by a single gunner.

Then there are the control programs needed to use nuclear dampers and meson screens.

A series of bay weapon programs, spinal mount programs, higher ratings for predict and maneuver/evade - that sort of thing.

Now that's what I'm getting at.

Still maintain GI is good for all turrets? What about batteries of turrets? Different programs for different weapon types, or just a GI for turret then another for bays and a third for spinal mounts?

Probably should be another for launch tubes.