Setting wind direction via NWI: only right-shifting allowed?

[CATIII]

Dear Pete,

here comes an interesting problem which is IMHO related to the NWI facility or to the weather processing code of FS2004 in general.

Weather setting and reading is done here from a separate host on the network running WideClient. For this we're using NWI through a self written Delphi GUI. Everything works great EXCEPT setting of wind direction in upper layers.

I'm always writing 3 wind layers [1..3] altogether with all necessary properties. After hours of testing I found the following phenomenon: If the wind direction of an upper layer (let's say no. 2) is numerically smaller than the previous one (no. 1), it gets screwed up and has (nearly!) nothing to do with the original one. Or to say it in other words: as long as the wind direction of the next higher level is greater than the one in the lower level, everything is fine. Winds can only shift to the right in the next higher level. Shifting to the left (smaller direction values) are not accepted by FS.

Below you see some results of tests that I made with FSInterrogate2. Of course I checked my software for probable mistakes, but the values which I send are identical to those which FSUIPC receives.

             FSUIPC Offsets   Test1   Test2   Test3   Test4   Test5   Test6
============================================================================
Wind Layer1: Written  (C902):  107     240     251     163     200     200
             Last Set (C502):  107     240     251     163     200     200
             Current  (C102):  107     240     251     163     200     200
============================================================================
Wind Layer2: Written  (C912):  115     270     189     133     110     270
             Last Set (C512):  115     270     189     133     110     270
             Current  (C112):  115     270     313!!   193!!   290!!   270
============================================================================
Wind Layer3: Written  (C922):  178     293     138     143      80      90
             Last Set (C522):  178     293     138     143      80      90
             Current  (C122):  178     293     240!!   143     140!!    90?!?!?

Tests 1 and 2 have no exclamation mark, they work fine since the wind direction in the next higher level is greater than the previous one. Tests 3, 4 and 5 clearly show the problem. The resulting values (which are current in FS9) are NOT identical to the written values. It seems as if FS takes the difference in direction between no.1 and 2 and just ADDS it to the no.1 direction. In Test 3: 251-189=62. 251+62=313. 313 degrees is what I get in no.2, but I want 189°. FS should take my absolutely supplied value and NOT just add the difference.

From my point of view I'm stuck in a blind alley now. There's no way for me to work-around this problem. Since you're the one who has the insight and experience with MS and FS, I'd like to ask if you know any solution for this dilemma. Let me know if you need more test results from me. I'm really interested in finding an acceptable solution.

Thank you very much for your help in advance.

Best regards,

CATIII

FS9.exe         9.1.0.40901
FSUIPC.dll      3.7.6.6
WideServer.dll  6.7.5.0
WideClient.exe  6.7.5.7
WeatherSet2.exe 1.4.0.0

PS: And last but not least: all smoothing/fading/changing facilities in FSUIPC are disabled.

[Pete Dowson]

here comes an interesting problem which is IMHO related to the NWI facility or to the weather processing code of FS2004 in general.

Yes. It's the FS "live" weather processing for wind layers (I think temperature layers have similar but less consequential problems). The same bug occurs in FS9 and FSX, and it doesn't just appliy to weather written through FSUIPC (or SimConnect in FSX), but also to FS's own downloaded weather.

Unfortunately, so far, there's no fix in FSX, so the 180 degree and similar windshifts are a really serious problem. The only way in FS9 to fix the matter is to use FSUIPC's wind smoothing, which actively overrides the resulting winds at the aircraft on its way from the WEATHER.DLL to the Simulation Engine (SIM1.DLL). I'm still looking for a way to do the same in FSX.

Regards

Pete