Realism- and gameplay-related hardcode fixes for SH3.EXE

[Yoriyn]

After testing new EnvSim.act on clear SH3 installation in campaign mode I can say:
- weather changing intervals for my XP/32bit is 1:33hour, so every one and a half hour weather change.

I don't see the point to making mess in game weather algorithms and change wind behavior. Weather behavior is different, and depends on your location on the map, season of the year and more. So we have different weather on North Atlantic and different on South or Mediterranean sea. Bad weather on North Atlantic is very common. I just finish reading a Jost Metzler book (one of the u-boat captain in WWII), and on all his three patrols on the North atlantic he had heavy storms.

In this moment stock game have weather changing intervals is 33-34hours. In reality weather change more often, almost all the time. So after 33h we have a chance to weather change, even if the wind do not change the fog, rain, clouds can. So changing weather intervals to 12-16h reduce fog duration to 12-16h instead 33-34h, without messing in algorithms.

If the fog change in second or third weather interval this give you:
33h interval - second interval after 66h (2,5days of constant fog) - third interval 99h ( more then 4days constant fog or rain etc)
12h intervals - secind interval 24h (give us only day with fog or rain) - in third interval 36h (1,5 day still ok for me)

[h.sie]

Thanks, Yoriyn, for first impressions. I like the watchcrew dancing on the bridge in TC=128x. I also tend to do as few changes as possible - only what is really necessary.

Let us wait and see what the others think. I for myself have no experience regarding weather because, as I mentioned multiple times, I never played a career.

I'm looking forward for a controverse discussion about weather with 1o people and 12 different opinions. .


Even if we decide to do some changes to windspeed, I'll make a special version for you with 12h Intervals - without any other change.
Done in 10 minutes.

[Stiebler]

@H.sie:

I've just sent you a PM, concerning Ollydbg2.

@Yoriyn:

H.sie has discovered that there is definitely a bug in the wind speed control. This needs to be fixed, since it may solve other weather problems too.

Stiebler.

[Yoriyn]

Stiebler: What weather bug you have on mind?

h.sie: Did you found any connection between the setting WeatherRndInterval= in campaign_xxx.mis and weather change intervals in game?

[h.sie]

@Yoriyin: No. There has been massive research in the past regarding these WeatherRndInterval settings, but with moderate/diffuse success.
So I think this part is somehow broken and not worth to be investigated furtheron. But maybe I'm wrong with this.

[Hitman]

Stiebler: What weather bug you have on mind?

Mate it's in the previous page ....

[Yoriyn]

Sorry, but I think I got it now little late but is better late then never

I don't think the bug is when program generate wind bigger then 15m/s. 15m/s it is modrate strong (moderate gale not a storm) wind on sea, so for me is strange the game wind scale finish on this 15. If you programming you exacly know what you want to do, so they must know the algorithm produce numbers from 0 to 30. It is more logic for me to produce winds between 0-30m/s then 0-15m/s, specialy for u-boats game, where most patrols have a place on north atlantic. For me this strange 15m/s barrier looks very, very strange like abandoned project or rush in developing.
Where are this hevy storms described by u-boat crews, where peolpe in the machine room was shooted in the "air" by waves. I think bigest problem is in weather change period (as I said before few days with fog), and another in wind 0m/s, which is imposible rare phenomena on north atlantic.

[h.sie]

@Yoriyn: You are right, but as Hitman & Stiebler said: One can enlarge the waves later in the scene.dat or using SH3-Commander.

The limitation to 0...15m/s is located on a different location in the code (at the very end) and so I also think this was a quick&dirty fix by the devs perhaps because with higher waves some unwanted side-effects occured.

I'll try a 2nd algorithm according to Stieblers idea: Damping down the output of the algorithm instead of delimiting. Or as a compromise, little damping, so that delimiting occurs very seldom.

I could also test what happens if I simply bypass the limit of 15 and allow higher speeds, but there surely will be side-effects....

[Yoriyn]

Only problem with bypass wind restriction can be CTD if not waves stay the same as are, because stock file scene.dat have a setting for a winds:

0-3m/s - small ribblets
4-7m/s - small waves
8-14m/s - bigger waves
15m/s or more - moderate waves

[Hitman]

The limitation to 0...15m/s is located on a different location in the code (at the very end) and so I also think this was a quick&dirty fix by the devs perhaps because with higher waves some unwanted side-effects occured.

It's difficult to tell why they would cap the wind speed at 15 knots. May be with higher limit the algorythm had a tendency to make weather even worse. May be they had no time to finish the job with the waves, but this seems rather improbable, since the "large" waves assigned to current 15 m/s wind are actually very large ones. In my opinion something did not work well with the algorythm, and they reduced windspeed to 15, then adapting the four sizes of waves to that range (0-15). Another telling element is the speed with which clouds move in the sky. With 15 m/s they are already rushing around, so it seems that indeed they simply adapted their maximum, originally planned fopr much harder winds, to 15 m/s and then scaled all the rest.

Only problem with bypass wind restriction can be CTD if not waves stay the same as are, because stock file scene.dat have a setting for a winds:

0-3m/s - small ribblets
4-7m/s - small waves
8-14m/s - bigger waves
15m/s or more - moderate waves

That shouldn't be a problem even if no bigger waves are added to the scene.dat. The reason is that the waves are defined only with START wind speed, i.e. *from* a certain wind speed, a certain wave model is used, until the wind reaches a strength where the scene.dat has a different wave, and then the new one is used. If you do not add different waves for windspeeds >15 m/s, then the game will simply keep using the current 15 m/s wave. The only difference you would notice is in the speed with which clouds move in the sky, unless other effects (reduce your speed due to aerodynamics, increase drag, etc) are programmed, which I do not know right now.

[Yoriyn]

I agree with you. We never find out if we do not try (of course if the h.sie give as a try ).

[h.sie]

So which factor should be used at the end to reduce weather change period?

[Yoriyn]

I'm voting something 12-14 hours.
I think is more realistic to have weather change in morning and evening the same day.

[h.sie]

that would be a factor of about 3 or 4.

time periods would also be reduced by the same factor.

[Yoriyn]

h.sie: can't understand your last post, what means factor 3 or 4?

Do not base only on my posts, mayby wait till others give they opinions. Together we can solve more problems and have different points of view, and knowledge.