New Weather Interface and Wind Direction

[JarangMandi]

Pete,

You're going to start dreading my posts soon, but...

I'm seeing a few problems with reading wind direction through the NWI with FS9 using FSUIPC 3.817. I've seen it manifest itself in all sorts of ways, and can't find a pattern to it, but the following steps will (should) reproduce it.

1. Fire up WeatherSet2

2. Fire up FS2004

3. Choose "Select a Flight" and choose the "Default Flight". For me this is still the default start at KSEA.

4. Select "Fly Now".

5. Enable FSUIPC Weather Logging.

6. Select weather theme "Clear Skies" from "World... Weather" dialog

7. Weatherset2 shows a single wind layer, speed and direction 0,0.

8. In FS9, select "user defined weather" from "World... Weather..." dialog, and then "advanced weather"

9. On the wind tab, insert a new wind layer, 25000ft, 120kts, 270 deg.

10. OK all the way out...

11. Weatherset2 shows two wind layers - layer 0 is unchanged; layer 1 is at 25000ft, 120kts BUT 090 deg.

The relevant lines from the log file read as follows...

   483946 >Change:  FS98 Pressure=0.0 mb
   483946 >Change:  FS98 Vis: range=0sm, (raw value=0)
   483946 >Change:  FS98 Dewpoint Control: 0.0C
   483946 >Change:  FS98 Precip Control: None, Intensity=0
   483946 >Change:  surface wind: to alt=2000ft AGL, dir=360T, vel=0.0, gust=0.0, turb=0, shear=0, var=0.0, Flags=0, AMSL=2380ft
   483946 >Change:  wind layer 0: to alt=25000ft AMSL, dir=90T, vel=120.0, gust=0.0, turb=0, shear=0, var=0.0
   483946 >Change:  Visibility[0]: range=60.0sm (96560m), from=0ft, to=15000ft
   483946  Results: Visibility[0]: range=60.0sm (96561m), from=0ft, to=15000ft
   483946  Results: FS98 Wind0: ground (430ft) to 1950ft AGL, dir 359M, vel 0, gust 0, turb 0
   483946  Results: FS98 Wind1: 2380ft to 25000ft AMSL, dir=89T, vel 120, gust 0, turb 0

I have taken a look at the raw data as reported with offset reads are logged and reads are executed from from offsets 0xC000 and 0xCC00, and the wind direction reported in there does match the value in the log (089deg) - but obviously not the value that was set (270deg). I've also tried with different direction values... an entry of 315 degrees comes back as 045, and an entry of 225 degrees comes back as 135 deg. My first impression was that it might be an overflow, but when I started changing the direction of the lower layer that didn't seem to hold true any more. When the direction on the lower layer is changed, all kinds of weird stuff happens!

I've not had the opportunity to try this on FSX yet as I'm working away from home and they won't give me a laptop that will run it! I'd be surprised if the problem was the same though as I understand that the underlying mechanisms are now totally different.

Do you see the same things happening?

Thanks (again) in advance,

Andy

[Pete Dowson]

You're going to start dreading my posts soon, but...

Well, not dread, but i must admit I cannot fathom what it is you are trying to prove. All of what you appear to be investigating is old hat. FS2004 weather and all of its ins and outs was sorted years ago. There's going to be no changes to any of the interfaces. They work quite well, better than FS's own facilities, and there's no more that can be done with the FS9 weather engine.

For that matter, there's going to be nothing further done for FSX either, but for different reasons. It needs to be sorted in SimConnect. I'm not about to go through the same hacking hell as I endured for FS9. The next change would be for ESP2 or FSXI, whichever is first.

I'm seeing a few problems with reading wind direction through the NWI with FS9 using FSUIPC 3.817.

You say that, but absolutely nothing in what then follows points to anything in the NWI wrong at all, only in the FS dialogues. You even reproduced a log extract showing the weather logged in other ways. All this is data extracted from FS.

You didn't even bother to check which was correct, either, did you? If you'd bothered to slew up to 25000 feet and use Shift+Z to see the weather actually at that altitude you would have found WeatherSet2 and the logging to be accurate (when you do this, bear in mind WeatherSet uses TRUE whilst the Shift+Z display is MAG).

All you are doing is proving, as others have before you, that FS's weather setting dialogues are pretty useless for setting weather very precisely. They are, in a word, buggy.

If you want to set weather reliably why not use WeatherSet2 for this as well? To do your experiment:

First slew up to 25000 feet and do Shift+Z twice to show the wind there. If you slew up after setting the wind it may well have changed by the time you get there owing to FS's dynamics.

Load WeatherSet2.

Clear all weather (Clear menu, "clear All").

Place cursor on the "0" under "Station" and Press Enter.

Type GLOB for the station name (or your airport if directly over one) and press Enter.

Place the cursor on the "1" under "Winds".

Press Insert.

Enter the wind data -- 25000 feet, 50 knots, 358 True, or whatever.

Press Enter

Check the readback in WeatherSet2 (degrees True) and that on the FS screen (degrees Mag).

What more do you want?

WeatherSet2 is 100% NWI. No other tricks. Everything it does you can do through that interface. It is supported and it works.

I've not had the opportunity to try this on FSX yet as I'm working away from home and they won't give me a laptop that will run it! I'd be surprised if the problem was the same though as I understand that the underlying mechanisms are now totally different.

No, they are not totally different. Most of the old bugs are there, and they added some more -- more complex ones. Also, because FSUIPC no longer hacks into FS code to control or read the weather, but relies on the promised facilities in SimConnect, it cannot set or read so much. Cloud types and thicknesses, for instance, are not directly controllable, and thicknesses impossible to read. The visibility is only indirectly manageable. The winds have a habit of multiplying spuriously in numbers of layers, and still exhibit near 180 degree shifts in direction. FSUIPC tries to overcome some of this, but doesn't entirely succeed -- which is why most of the weather filtering facilities in FSUIPC4 are defaulted off, and only provided as options with health warnings attached.

ASX also does its best with FSX, and, because it is a weather setting program, unlike FSUIPC (which tries to be a weather assisting program), ASX is rather more successful in achieving its aims. But it has been a great deal of trouble to do for its author, and it still isn't 100% satisfactory, only the best likely to be achieved at present.

BTW if you want support here to continue will you kindly update your FSUIPC. Versions 3.81X have not been supported since 3.82 came out in July, and the latest interim update to that is 3.845. So, 3.820 is the earliest you can be using for any form of support. Please do read the relevant Announcements above which tell you these things.

Regards

Pete

[JarangMandi]

Well, not dread, but i must admit I cannot fathom what it is you are trying to prove. All of what you appear to be investigating is old hat. FS2004 weather and all of its ins and outs was sorted years ago.

I'm not trying to prove anything - just to understand FS weather and its interaction through the NWI to FSUIPC. I'm starting to develop software against FS9 later than most, and the searches I ran through this forum didn't turn up any answers, although that might have been down to the search terms I was using.

You say that, but absolutely nothing in what then follows points to anything in the NWI wrong at all, only in the FS dialogues. You even reproduced a log extract showing the weather logged in other ways. All this is data extracted from FS.

You didn't even bother to check which was correct, either, did you? If you'd bothered to slew up to 25000 feet and use Shift+Z to see the weather actually at that altitude you would have found WeatherSet2 and the logging to be accurate (when you do this, bear in mind WeatherSet uses TRUE whilst the Shift+Z display is MAG).

That information answers the question I was asking. You're right - I didn't check the winds aloft in FS9 by actually going there and I apologise for not doing this before posting.

Type GLOB for the station name (or your airport if directly over one) and press Enter.

Should I expect to see "GLOB" in the data that WeatherSet2 displays after doing this? I don't - and I've noticed the same behaviour when trying to read weather for "GLOB" from code; I always get a zeroed string in chICAO instead - which I understand to mean the weather at the aircraft location.

BTW if you want support here to continue will you kindly update your FSUIPC. Versions 3.81X have not been supported since 3.82 came out in July, and the latest interim update to that is 3.845. So, 3.820 is the earliest you can be using for any form of support.

The information regarding version number was drawn from the FSUIPC dialog - the date in that dialog is July 16th. This doesn't match the version of the file that appears in the file properties when examined through Windows Explorer which is indeed 3.82. I've checked that it's this file that's being used by removing it from the "modules" directory.

[Pete Dowson]

Should I expect to see "GLOB" in the data that WeatherSet2 displays after doing this? I don't - and I've noticed the same behaviour when trying to read weather for "GLOB" from code; I always get a zeroed string in chICAO instead - which I understand to mean the weather at the aircraft location.

Yeah, it should show GLOB. It does in FSX. Some sort of timing change in FSUIPC3 I think. I'm not messing with it now unless it turns out to be a problem, as it seems to have been like that for a good couple of years. If it allows you to set weather, it's on "GLOB" and if not it's only weather at aircraft.

The information regarding version number was drawn from the FSUIPC dialog - the date in that dialog is July 16th. This doesn't match the version of the file that appears in the file properties when examined through Windows Explorer which is indeed 3.82. I've checked that it's this file that's being used by removing it from the "modules" directory.

Oh, right. Yes, sorry, I forgot that the final 3.81x interim was released as 3.82 and everything was changed except I forgot to update the dialogue. Apologies. It was fixed in an interim update to 3.821 immediately after.

Incidentally, you can also see the correct version number at the top of the Log file.

Please use 3.845 then I won't be confused! ;-)

Regards

Pete