Which of the files contain aircraft?

[el cid again]

ORIGINAL: Dili

It has been said the vanilla maneuver ratings are a combination of speed, ROC, turning, God knows what all, but if so, all of this fruit salad was not necessary to arrive at the "magic number", since ROC and speed are already calculated seperately.

My point entirely. It can only be taken down if ROC and Speed have a much more restricted propose and dont affect air combat. We need to know that.

Think we're creeping up on it.

JWE if we dont know how air combat model works in witp the end result can be  gigantic, wonderful and wrong.

El Cid since you seems to have contacts in Matrix what about you bug them concerning air combat model?

The problem is several fold:

a) Priorities. The air model works well. Other problems are worse - so they get higher priority. This is for me as well as for Matrix. I have put a lot of effort into the air model already - and before much more can be justified - we need to figure out how badly (or well) it is working in the modified form?

b) Costs. Time and (for Matrix) money. Relate to priority above.

c) Structure. There is a limit to what can be expected from the approach being used. A fundamental rewrite is required to make a fundamental change. [OTH I can change data at no cost whatever to Matrix] So that means that we cannot expect a fundamental rewrite unless a fundamental commitment is made to rework a major code routine - as might be the case with a WITP II.

d) Identification: major issues ARE identified and Matrix will (or will not) address them as time/resources permits. For example, there is a known issue if more than 50 planes are on a side. The problem is wholly understood. When it is addressed is up to Matrix.

Note that if you work out air combat by a very sophisticated model, by hand, you will get very similar results to WITP in many cases. Particularly for smaller combats (under 50 planes per side). Also note that the model is inherently better than we presently see: IF we give it better data on why plane X is better than plane Y - it will use it. So my focus has been on getting better RELATIVE data - since I cannot change the model itself. Once we have that as good as we can do - we can better assess what might be the most important things to change in the model? the 50 planes per side - and the ammunition limit (that is NO ammunition limit) - probably are going to stay at the top of the list - so I think the grasp of Matrix and the Forum is well formed re air combat.

I repeat: a professional programmer and game developer I know says this is the best air model ever developed. He probably is not right - but in this cost range - he probably IS right. It is amazing how much it does well. We will also never be satisfied - however much better they make it - we will always want more. Matrix lives in the real world of cost/return analysis: they need to figure out what can be justified?

[Dili]

I didnt talk about code, exe etc... When i asked you were not able to tell with certainity if the Climb rate and Speed field is used for Air Combat. We need that info.

[el cid again]

ORIGINAL: Dili

I didnt talk about code, exe etc... When i asked you were not able to tell with certainity if the Climb rate and Speed field is used for Air Combat. We need that info.

I was told that the immediate air combat routine was dominated by the Maneuverability value. It is unlikely anyone knows every routine that uses every value - and an attempt to figure that out would probably require looking at a lot of different routines. On a different question a few days ago, I asked the difference between this setting and that one, and was told there were probably hundreds of instances the setting was used, and the answer was unknown. It turned out to be easier for me to look for an answer without the proper theoretical understanding, and I found it a few hours later.
I am an old computer person, from the era of front panels, program (and operating system) listings, where NOTHING was hidden. I would prefer total information to work from - and instant and automatic register dumps at all times. The modern world does not work that way - and most of the time - even if you can see the listing - you cannot understand it in a reasonable (or even unreasonable) time frame. This is the price we pay for our very sophisticated and complex machines, operating systems and programs in this age. People who can do analysis of a problem - test people like me - are very often as good as you are going to get - for better or for worse. Having been told what the routine uses by the specialist who has a better grasp than anyone else (probably) - I work in the context of that information. I dream of and wish for more - but it is impossible we will get everything. Proposals to make WITP I open source seem to have been tabled as well. And I consider it unethical to reverse engineer and reconstruct the code, having agreed as a condition of purchase not to do that, so I won't try. Ultimately Matrix has served our community not only well, but better than any other company ever has done. We need to respect Matrix interests and choices.

[m10bob]

ORIGINAL: el cid again

ORIGINAL: Dili

It has been said the vanilla maneuver ratings are a combination of speed, ROC, turning, God knows what all, but if so, all of this fruit salad was not necessary to arrive at the "magic number", since ROC and speed are already calculated seperately.

My point entirely. It can only be taken down if ROC and Speed have a much more restricted propose and dont affect air combat. We need to know that.

Think we're creeping up on it.

JWE if we dont know how air combat model works in witp the end result can be  gigantic, wonderful and wrong.

El Cid since you seems to have contacts in Matrix what about you bug them concerning air combat model?

The problem is several fold:

a) Priorities. The air model works well. Other problems are worse - so they get higher priority. This is for me as well as for Matrix. I have put a lot of effort into the air model already - and before much more can be justified - we need to figure out how badly (or well) it is working in the modified form?

b) Costs. Time and (for Matrix) money. Relate to priority above.

c) Structure. There is a limit to what can be expected from the approach being used. A fundamental rewrite is required to make a fundamental change. [OTH I can change data at no cost whatever to Matrix] So that means that we cannot expect a fundamental rewrite unless a fundamental commitment is made to rework a major code routine - as might be the case with a WITP II.

d) Identification: major issues ARE identified and Matrix will (or will not) address them as time/resources permits. For example, there is a known issue if more than 50 planes are on a side. The problem is wholly understood. When it is addressed is up to Matrix.

Note that if you work out air combat by a very sophisticated model, by hand, you will get very similar results to WITP in many cases. Particularly for smaller combats (under 50 planes per side). Also note that the model is inherently better than we presently see: IF we give it better data on why plane X is better than plane Y - it will use it. So my focus has been on getting better RELATIVE data - since I cannot change the model itself. Once we have that as good as we can do - we can better assess what might be the most important things to change in the model? the 50 planes per side - and the ammunition limit (that is NO ammunition limit) - probably are going to stay at the top of the list - so I think the grasp of Matrix and the Forum is well formed re air combat.

I repeat: a professional programmer and game developer I know says this is the best air model ever developed. He probably is not right - but in this cost range - he probably IS right. It is amazing how much it does well. We will also never be satisfied - however much better they make it - we will always want more. Matrix lives in the real world of cost/return analysis: they need to figure out what can be justified?

And all of this is exactly why I mentioned the law of diminishing returns, especially at this scale!

I feel if nothing more than ROC and ceiling are made accurate, it will be a grand improvement.
Too, the look at the empty/payload ranges will be very revealing.

[witpqs]

ORIGINAL: m10bob
Too, the look at the empty/payload ranges will be very revealing.

Big time.

[JWE]

ORIGINAL: Dili

JWE if we dont know how air combat model works in witp the end result can be  gigantic, wonderful and wrong.

Too right! But as I mentioned to Bob, somewhere up there, this is to satisfy my curiousity and to see if a VPP can't do some air predictions. I do know how the air combat model(s) work, so that shouldn't be an issue. Besides, I normalize to the corners, and this isn't for RHS in any case.

Having a lot of fun though, finding a lot of conditional relationships; BTW, found 4 different ways speed effects maneuver. Ciao.

[m10bob]

Even with a "uniform" reference work the planes themselves had some wonderful anomolies.
For instance, the P 39 had an initial ROC of nearly 4000 ft/min, (very close to the A6m2), but it ran out of steam quickly and the sustained average ROC was only about 2450 ft/min..
I intend to send my cleaned-up mod to Sid later today for his inspection, but I suspect he already has it, and possibly then some.........


BTW...NONE of these ROC nor ceiling changes should throw the game akilter since I have found maybe 1/10th of them have identical figures to the source book I am using....
This same book also gives GREAT justification for changes in range.
For now, I am not touching that,yet...

[Dili]

I think it is better to extrapolate ROC from time to 5000m or 6000m or any other altitude than just ROC numbers.

[m10bob]

I suspect that in-game at this scale, it provided a modifier to a fighter planes' ability to intercept an enemy bomber unit coming in at a different altitude.
If so, this would be very accurate, especially at the scale of the game.