Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
Thanks Tim! On Wed, 15 Aug 2007, Reagan Thomas wrote: It looks like this fix needs to be applied to the weather interpolation in the plib branch also. Yes, it is needed also in the plib branch. I have tested the patch there too. Cheers, Anders -- --- Anders Gidenstam mail: anders(at)gidenstam.org WWW: http://www.gidenstam.org/FlightGear/JSBSim-LTA/ - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Committed; thanks. Tim Anders Gidenstam wrote: On Sun, 12 Aug 2007, Hans Fugal wrote: That's because there is no custom comparison predicate for comparing bucket*. vectorbucket* is just a list of pointers, and so uses the sort order for pointers. You'd have to set up an STL functor to sort them, which is ugly and difficult (but certainly doable). If sorting is unnecessary it's a better way to go, obviously. :-) Yep, that's what I figured out eventually. I think it is safer to keep the sorting for now - in case e.g. the user changes the layer elevations using the property browser. The patch included adds a comparison predicate for buckets. Please test it and commit it if it is acceptable. It works fine here. Cheers, Anders diff --git a/src/Environment/environment_ctrl.cxx b/src/Environment/environment_ctrl.cxx index a1ed5f2..0ca2cba 100644 --- a/src/Environment/environment_ctrl.cxx +++ b/src/Environment/environment_ctrl.cxx @@ -231,7 +231,7 @@ FGInterpolateEnvironmentCtrl::read_table (const SGPropertyNode * node, table.push_back(b); } } -sort(table.begin(), table.end()); +sort(table.begin(), table.end(), bucket::lessThan); } void @@ -312,6 +312,11 @@ FGInterpolateEnvironmentCtrl::bucket::operator (const bucket b) const return (altitude_ft b.altitude_ft); } +bool +FGInterpolateEnvironmentCtrl::bucket::lessThan(bucket *a, bucket *b) +{ +return (a-altitude_ft) (b-altitude_ft); +} diff --git a/src/Environment/environment_ctrl.hxx b/src/Environment/environment_ctrl.hxx index f561f05..cd01b55 100644 --- a/src/Environment/environment_ctrl.hxx +++ b/src/Environment/environment_ctrl.hxx @@ -134,6 +134,8 @@ private: double altitude_ft; FGEnvironment environment; bool operator (const bucket b) const; +// LessThan predicate for bucket pointers. +static bool lessThan(bucket *a, bucket *b); }; void read_table (const SGPropertyNode * node, vectorbucket * table); - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel - -- Red Hat France SARL, 171 Avenue Georges Clemenceau 92024 Nanterre Cedex, France. Siret n° 421 199 464 00056 -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.7 (GNU/Linux) Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org iD8DBQFGwxp3eDhWHdXrDRURAi06AJ9HC5lPWthjDXoBhb7C1WaQlLwfGACgph1s FgxdcvBCmHOD/yjfJjMpLVQ= =0m5Y -END PGP SIGNATURE- - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
Tim Moore wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Committed; thanks. Tim It looks like this fix needs to be applied to the weather interpolation in the plib branch also. Anders Gidenstam wrote: On Sun, 12 Aug 2007, Hans Fugal wrote: That's because there is no custom comparison predicate for comparing bucket*. vectorbucket* is just a list of pointers, and so uses the sort order for pointers. You'd have to set up an STL functor to sort them, which is ugly and difficult (but certainly doable). If sorting is unnecessary it's a better way to go, obviously. :-) Yep, that's what I figured out eventually. I think it is safer to keep the sorting for now - in case e.g. the user changes the layer elevations using the property browser. The patch included adds a comparison predicate for buckets. Please test it and commit it if it is acceptable. It works fine here. Cheers, Anders diff --git a/src/Environment/environment_ctrl.cxx b/src/Environment/environment_ctrl.cxx index a1ed5f2..0ca2cba 100644 --- a/src/Environment/environment_ctrl.cxx +++ b/src/Environment/environment_ctrl.cxx @@ -231,7 +231,7 @@ FGInterpolateEnvironmentCtrl::read_table (const SGPropertyNode * node, table.push_back(b); } } -sort(table.begin(), table.end()); +sort(table.begin(), table.end(), bucket::lessThan); } void @@ -312,6 +312,11 @@ FGInterpolateEnvironmentCtrl::bucket::operator (const bucket b) const return (altitude_ft b.altitude_ft); } +bool +FGInterpolateEnvironmentCtrl::bucket::lessThan(bucket *a, bucket *b) +{ +return (a-altitude_ft) (b-altitude_ft); +} diff --git a/src/Environment/environment_ctrl.hxx b/src/Environment/environment_ctrl.hxx index f561f05..cd01b55 100644 --- a/src/Environment/environment_ctrl.hxx +++ b/src/Environment/environment_ctrl.hxx @@ -134,6 +134,8 @@ private: double altitude_ft; FGEnvironment environment; bool operator (const bucket b) const; +// LessThan predicate for bucket pointers. +static bool lessThan(bucket *a, bucket *b); }; void read_table (const SGPropertyNode * node, vectorbucket * table); - -- Red Hat France SARL, 171 Avenue Georges Clemenceau 92024 Nanterre Cedex, France. Siret n° 421 199 464 00056 -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.7 (GNU/Linux) Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org iD8DBQFGwxp3eDhWHdXrDRURAi06AJ9HC5lPWthjDXoBhb7C1WaQlLwfGACgph1s FgxdcvBCmHOD/yjfJjMpLVQ= =0m5Y -END PGP SIGNATURE- -- Reagan Thomas - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
--- Georg Vollnhals wrote: I applied your patch and just made a flight from EDDV to EDDW (with --enable-real-weather-fetch) through several weather-zones (which I could recognize by the changing cloud display). First you cured the faulty behaviour with your patch :-). And the weather interpolation seems to work very fine. I could only identify the change between different METAR zones by the changing cloud textures - the weather influence on the aircraft was very smooth (no sudden up-down anymore) and I could hold my flightlevel by just changing the trim over some time. Great, I'm glad we've got to the bottom of this. I've tested Anders' patch on my system as well, and it works fine. Could someone please commit it? I think the layer interpolation code (which is quite separate from my METAR interpolation code) could do with a bit more attention. As well as the sort() bug which has been present for ages, Anders and I spotted a couple of other areas where there was suspect code. In particular, the boundary cases when determining which environments to interpolate look suspect. -Stuart ___ Yahoo! Answers - Got a question? Someone out there knows the answer. Try it now. http://uk.answers.yahoo.com/ - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
On Wed, 8 Aug 2007, Stuart Buchanan wrote: I'd be very grateful if you could help me diagnose the problem, as I did quite a lot of testing, but didn't encounter this myself. One possibility is that my interpolation is fighting with something else trying to write to the weather properties. Hi! With Stuart's help I've looked closer at this and I think I've tracked down the cause of the problem: At least on my computer the sort() call on line 234 in Environment/environment_ctrl.cxx sorts the vector entries by memory address instead of altitude, i.e. the custom comparison predicate is not used. This causes the tables of environment conditions to be reordered into a wrong order at some weather updates, depending, basically, on where the memory allocator places the objects. (Btw. why are they are freshly allocated for each update?) The badly ordered tables cause the interpolation of the local weather to start from the wrong conditions and, hence, the result is quite different from the weather before the weather update - a wall of weather. How to fix this? The obvious way would be to ensure that the proper ordering predicate is used by sort(). OTOH I'm not sure the sort is needed at all - the entries are read from the property three, where I think they are already sorted. I'm testing with out the sort now, and so far I have seen no problems. Cheers, Anders -- --- Anders Gidenstam mail: anders(at)gidenstam.org WWW: http://www.gidenstam.org/FlightGear/JSBSim-LTA/ - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
On 8/12/07, Anders Gidenstam [EMAIL PROTECTED] wrote: On Wed, 8 Aug 2007, Stuart Buchanan wrote: I'd be very grateful if you could help me diagnose the problem, as I did quite a lot of testing, but didn't encounter this myself. One possibility is that my interpolation is fighting with something else trying to write to the weather properties. Hi! With Stuart's help I've looked closer at this and I think I've tracked down the cause of the problem: At least on my computer the sort() call on line 234 in Environment/environment_ctrl.cxx sorts the vector entries by memory address instead of altitude, i.e. the custom comparison predicate is not used. This causes the tables of environment conditions to be reordered into a wrong order at some weather updates, depending, basically, on where the memory allocator places the objects. (Btw. why are they are freshly allocated for each update?) That's because there is no custom comparison predicate for comparing bucket*. vectorbucket* is just a list of pointers, and so uses the sort order for pointers. You'd have to set up an STL functor to sort them, which is ugly and difficult (but certainly doable). If sorting is unnecessary it's a better way to go, obviously. :-) When I comment out the sort, it works well for me. -- Hans Fugal Fugal Computing - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
On Sunday 12 August 2007 13:48, Anders Gidenstam wrote: On Wed, 8 Aug 2007, Stuart Buchanan wrote: I'd be very grateful if you could help me diagnose the problem, as I did quite a lot of testing, but didn't encounter this myself. One possibility is that my interpolation is fighting with something else trying to write to the weather properties. Hi! With Stuart's help I've looked closer at this and I think I've tracked down the cause of the problem: At least on my computer the sort() call on line 234 in Environment/environment_ctrl.cxx sorts the vector entries by memory address instead of altitude, i.e. the custom comparison predicate is not used. This causes the tables of environment conditions to be reordered into a wrong order at some weather updates, depending, basically, on where the memory allocator places the objects. (Btw. why are they are freshly allocated for each update?) The badly ordered tables cause the interpolation of the local weather to start from the wrong conditions and, hence, the result is quite different from the weather before the weather update - a wall of weather. How to fix this? The obvious way would be to ensure that the proper ordering predicate is used by sort(). OTOH I'm not sure the sort is needed at all - the entries are read from the property three, where I think they are already sorted. I'm testing with out the sort now, and so far I have seen no problems. Cheers, Anders I wonder if this might be the cause of a bug I used to see where the weather settings i.e. wind speed direction for different altitudes were not being applied correctly. For example, sitting on the runway I should have seen a wind speed of 3 kts but often I would actually get 6 kts, which was the setting for 500 ft. Then once I'd taken off and got up to 3000ft, instead of seeing the correct 10 kts I'd often get either 20 or 30 kts, which were the settings for 6000ft and 9000ft respectively. Re-applying the weather settings would then just randomly select one of the levels (3000ft, 6000ft or 9000ft). For example, as I re-applied the weather settings (without actually changing anything) I would see the wind speed/direction jump at random between the 3000ft, 6000ft 9000ft settings. LeeE - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
On Sun, 12 Aug 2007, Hans Fugal wrote: That's because there is no custom comparison predicate for comparing bucket*. vectorbucket* is just a list of pointers, and so uses the sort order for pointers. You'd have to set up an STL functor to sort them, which is ugly and difficult (but certainly doable). If sorting is unnecessary it's a better way to go, obviously. :-) Yep, that's what I figured out eventually. I think it is safer to keep the sorting for now - in case e.g. the user changes the layer elevations using the property browser. The patch included adds a comparison predicate for buckets. Please test it and commit it if it is acceptable. It works fine here. Cheers, Anders -- --- Anders Gidenstam mail: anders(at)gidenstam.org WWW: http://www.gidenstam.org/FlightGear/JSBSim-LTA/diff --git a/src/Environment/environment_ctrl.cxx b/src/Environment/environment_ctrl.cxx index a1ed5f2..0ca2cba 100644 --- a/src/Environment/environment_ctrl.cxx +++ b/src/Environment/environment_ctrl.cxx @@ -231,7 +231,7 @@ FGInterpolateEnvironmentCtrl::read_table (const SGPropertyNode * node, table.push_back(b); } } -sort(table.begin(), table.end()); +sort(table.begin(), table.end(), bucket::lessThan); } void @@ -312,6 +312,11 @@ FGInterpolateEnvironmentCtrl::bucket::operator (const bucket b) const return (altitude_ft b.altitude_ft); } +bool +FGInterpolateEnvironmentCtrl::bucket::lessThan(bucket *a, bucket *b) +{ +return (a-altitude_ft) (b-altitude_ft); +} diff --git a/src/Environment/environment_ctrl.hxx b/src/Environment/environment_ctrl.hxx index f561f05..cd01b55 100644 --- a/src/Environment/environment_ctrl.hxx +++ b/src/Environment/environment_ctrl.hxx @@ -134,6 +134,8 @@ private: double altitude_ft; FGEnvironment environment; bool operator (const bucket b) const; +// LessThan predicate for bucket pointers. +static bool lessThan(bucket *a, bucket *b); }; void read_table (const SGPropertyNode * node, vectorbucket * table); - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
Anders Gidenstam schrieb: On Sun, 12 Aug 2007, Hans Fugal wrote: That's because there is no custom comparison predicate for comparing bucket*. vectorbucket* is just a list of pointers, and so uses the sort order for pointers. You'd have to set up an STL functor to sort them, which is ugly and difficult (but certainly doable). If sorting is unnecessary it's a better way to go, obviously. :-) Yep, that's what I figured out eventually. I think it is safer to keep the sorting for now - in case e.g. the user changes the layer elevations using the property browser. The patch included adds a comparison predicate for buckets. Please test it and commit it if it is acceptable. It works fine here. Cheers, Anders Hej Anders, tack so mycket! I applied your patch and just made a flight from EDDV to EDDW (with --enable-real-weather-fetch) through several weather-zones (which I could recognize by the changing cloud display). First you cured the faulty behaviour with your patch :-). And the weather interpolation seems to work very fine. I could only identify the change between different METAR zones by the changing cloud textures - the weather influence on the aircraft was very smooth (no sudden up-down anymore) and I could hold my flightlevel by just changing the trim over some time. It seems FlightGear has made some further step to more realism with this weather interpolation method. A big thank you to all who made this possible, to the author (Stuart Buchanan), Anders Gidenstam for the patch and all other who helped :-) Regards Georg EDDW - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
Hi Hans, Georg, --- Hans Fugal wrote: Hi Stuart, On 8/8/07, Stuart Buchanan wrote: I'd be very grateful if you could help me diagnose the problem, as I did quite a lot of testing, but didn't encounter this myself. Happy to help. One possibility is that my interpolation is fighting with something else trying to write to the weather properties. Thanks very much for the log files from you both. From what I can see, the interpolation itself looks fine, except that that it is continually writing the final wind values. I think there must be something fighting the interpolation, possibly to add some variation. However, as it isn't writing to the property itself (otherwise we'd see the other value being written), I wonder if it is the function that is tied to the get of the property? Re-reading the code, I think continual writing of the wind values might represent a small bug in that the inteprolation code doesn't first check if the wind is already correct, but I wouldn't expect that to result in any osscillations. Hans's log showed the wind varying (27012KT 240V310), which I initially thought might be related, but Georg's log doesn't show this, so I think that is probably a red herring. I will investigate further myself. Unfortunately I am away from my FG computer at the moment, so I may not manage to look at it in more detail until the weekend. Thanks -Stuart ___ Yahoo! Mail is the world's favourite email. Don't settle for less, sign up for your free account today http://uk.rd.yahoo.com/evt=44106/*http://uk.docs.yahoo.com/mail/winter07.html - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
Hi Stuart, I saw the variable wind in the METAR after I had posted, it is indeed a red herring. It happens (as you can see with Georg's post) no matter what the wind. We have some more information from IRC. The properties you asked for don't change, as you're aware, but some other properties exhibit interesting behaviors. /environment/wind-from-heading-deg is the one flopping back and forth from e.g. 270 and 280. This corresponds to the windsock dance. /environment/config/boundary/entry/wind-from-heading-deg stays at 270. If you change this value, it ramps back to 270 smoothly, which I assume is the intended behavior. /environment/config/boundary/entry[1]/wind-from-heading-deg stays at 280 (10 degrees difference). If you change this one, it snaps back immediately. If you change entry[0], then this one changes as well to maintain the 10 degrees separation, and they both ramp together back to 270/280. As you can see, /environment/wind-from-heading-deg flips back and forth between the values of the boundary entries. Hopefully that gives you a clue as to what's going on. Thanks! On 8/9/07, Stuart Buchanan [EMAIL PROTECTED] wrote: Hi Hans, Georg, --- Hans Fugal wrote: Hi Stuart, On 8/8/07, Stuart Buchanan wrote: I'd be very grateful if you could help me diagnose the problem, as I did quite a lot of testing, but didn't encounter this myself. Happy to help. One possibility is that my interpolation is fighting with something else trying to write to the weather properties. Thanks very much for the log files from you both. From what I can see, the interpolation itself looks fine, except that that it is continually writing the final wind values. I think there must be something fighting the interpolation, possibly to add some variation. However, as it isn't writing to the property itself (otherwise we'd see the other value being written), I wonder if it is the function that is tied to the get of the property? Re-reading the code, I think continual writing of the wind values might represent a small bug in that the inteprolation code doesn't first check if the wind is already correct, but I wouldn't expect that to result in any osscillations. Hans's log showed the wind varying (27012KT 240V310), which I initially thought might be related, but Georg's log doesn't show this, so I think that is probably a red herring. I will investigate further myself. Unfortunately I am away from my FG computer at the moment, so I may not manage to look at it in more detail until the weekend. Thanks -Stuart ___ Yahoo! Mail is the world's favourite email. Don't settle for less, sign up for your free account today http://uk.rd.yahoo.com/evt=44106/*http://uk.docs.yahoo.com/mail/winter07.html - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- Hans Fugal Fugal Computing - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
Hi All, --- Hans Fugal wrote: I did some more investigating, and discovered that it is still happening even fair weather, it's just that the wind speed is low enough that the effect isn't very pronounced. Also, taxi up to a windsock and watch the effect. I'm very sorry you are seeing problems related to my patch. Thanks very much to all who have reported it. I'd be very grateful if you could help me diagnose the problem, as I did quite a lot of testing, but didn't encounter this myself. One possibility is that my interpolation is fighting with something else trying to write to the weather properties. Could you explain what you are seeing in terms of wind direction and strength? Is it: - oscillating in direction and/or speed - rotating continually? - chaotic? Are you seeing the problem generally, or only when the METAR changes as you fly from one zone to another? If it is easily reproducable, could you run with the following command-line parameters and post the results, either to the list or directly to me: --trace-write=/environment/metar/real-metar --trace-write=/environment/metar/base-wind-range-from --trace-write=/environment/metar/base-wind-range-to --trace-write=/environment/metar/base-wind-speed-kt --trace-write=/environment/metar/gust-wind-speed-kt --trace-write=/environment/config/boundary/entry/wind-speed-kt --trace-write=/environment/config/boundary/entry/wind-from-heading-deg If you could also describe the specific flight that repros the problem that would be most helpful. Incidentally, I was trying to add a layer of clouds with --enable-real-weather-fetch. No matter what weather scenario I had chosen, adding clouds fails to have any effect (and when opening the clouds dialog again, everything is reset to -). I don't know if this is related, but it doesn't seem very desirable to have the weather dialogs do nothing useful when real-weather-fetch is enabled. This behaviour was present before my patch went in. Whenever a new METAR is read, the cloud config is reset. So, even though previously you could change the current cloud config, your changes would be lost when the next METAR was read - 15 minutes if the same station, within a minute if you moved to a new station. My patch effectively reduces these periods to 0. To set the clouds manually, I think you'd need to set the scenario to none, but see below. It might be a good idea to modify the dialog slightly to disable the configuration if METAR is active, along with a short explanatory message. Semantically, am I right that for weather scenarios, METAR is the real weather, Thunderstorm is thunderstorm-like weather (no relation to real weather?), fair is easy flying (again, no relation to real weather?), and none means no scenario (manual control?). That's what I think they should mean but I'm not convinced that that is what they mean (or anything else that would make sense). That's the theory. However, as I recall, the GUI doesn't select the correct scenario initially, so you may be running on METAR, open the dialog and fine none selected. However, if you click Apply, it should change the scenario to none properly. Once again, my apologies if this has spoilt anyone's flights. If this is causing people problems, it should probably be backed out. However, I would like to get to the bottom of the issue, as a (working) patch to interpolate METAR is something that we need. -Stuart ___ Yahoo! Mail is the world's favourite email. Don't settle for less, sign up for your free account today http://uk.rd.yahoo.com/evt=44106/*http://uk.docs.yahoo.com/mail/winter07.html - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
On Wed, 8 Aug 2007, Stuart Buchanan wrote: I'd be very grateful if you could help me diagnose the problem, as I did quite a lot of testing, but didn't encounter this myself. One possibility is that my interpolation is fighting with something else trying to write to the weather properties. Could you explain what you are seeing in terms of wind direction and strength? Is it: - oscillating in direction and/or speed - rotating continually? - chaotic? Hi, I also experience this problem. For me it seems the weather flips between two states, e.g. wind direction either 260 or 270 degrees and two different wind strengths. Are you seeing the problem generally, or only when the METAR changes as you fly from one zone to another? I see it pretty much all the time. In particular, I see it at KSFO runway 28L and 28R. If it is easily reproducable, could you run with the following command-line parameters and post the results, either to the list or directly to me: --trace-write=/environment/metar/real-metar --trace-write=/environment/metar/base-wind-range-from --trace-write=/environment/metar/base-wind-range-to --trace-write=/environment/metar/base-wind-speed-kt --trace-write=/environment/metar/gust-wind-speed-kt --trace-write=/environment/config/boundary/entry/wind-speed-kt --trace-write=/environment/config/boundary/entry/wind-from-heading-deg I'll try this tonight. Cheers, Anders -- --- Anders Gidenstam mail: anders(at)gidenstam.org WWW: http://www.gidenstam.org/FlightGear/JSBSim-LTA/ - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
Hans Fugal wrote: Semantically, am I right that for weather scenarios, METAR is the real weather, Thunderstorm is thunderstorm-like weather (no relation to real weather?), fair is easy flying (again, no relation to real weather?), and none means no scenario (manual control?). That's what I think they should mean but I'm not convinced that that is what they mean (or anything else that would make sense). This is right, except that in the 'none' scenario the weather is still updated by the metar if metar is enabled. I can not remember if this is a bug or a feature. HJ. - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
On 8/8/07, Harald JOHNSEN [EMAIL PROTECTED] wrote: This is right, except that in the 'none' scenario the weather is still updated by the metar if metar is enabled. I can not remember if this is a bug or a feature. If I may vote, I call it a bug. If there is no difference between none and METAR then one should be removed. I'd like to be able to set weather from time to time, and ignore METAR, without typing --disable-real-weather-fetch at the command line (I have the enable line in my .fgfsrc). - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
Hi Stuart, On 8/8/07, Stuart Buchanan [EMAIL PROTECTED] wrote: I'd be very grateful if you could help me diagnose the problem, as I did quite a lot of testing, but didn't encounter this myself. Happy to help. One possibility is that my interpolation is fighting with something else trying to write to the weather properties. Could you explain what you are seeing in terms of wind direction and strength? Is it: - oscillating in direction and/or speed - rotating continually? - chaotic? Definitely oscillating between two direction/speed settings (looking at the windsock makes this clear). Timing is not as regular, but roughly on the order of 1/2 to 1 second between changes. The problem is continual and omnipresent. If it is easily reproducable, could you run with the following command-line parameters and post the results, either to the list or directly to me: --trace-write=/environment/metar/real-metar --trace-write=/environment/metar/base-wind-range-from --trace-write=/environment/metar/base-wind-range-to --trace-write=/environment/metar/base-wind-speed-kt --trace-write=/environment/metar/gust-wind-speed-kt --trace-write=/environment/config/boundary/entry/wind-speed-kt --trace-write=/environment/config/boundary/entry/wind-from-heading-deg Attached Once again, my apologies if this has spoilt anyone's flights. If this is causing people problems, it should probably be backed out. However, I would like to get to the bottom of the issue, as a (working) patch to interpolate METAR is something that we need. I'd like to see it fixed rather than backed out, but whatever is best. It doesn't make things unflyable (usually), just unpleasant. log Description: Binary data - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
Could you explain what you are seeing in terms of wind direction and strength? Is it: - oscillating in direction and/or speed - rotating continually? - chaotic? Are you seeing the problem generally, or only when the METAR changes as you fly from one zone to another? Generally, directly after starting from the airport. But - as already mentioned in a former mail - I have to set --enable-real-weather-fetch. Without this there are no problems visible. If it is easily reproducable, could you run with the following command-line parameters and post the results, either to the list or directly to me: --trace-write=/environment/metar/real-metar --trace-write=/environment/metar/base-wind-range-from --trace-write=/environment/metar/base-wind-range-to --trace-write=/environment/metar/base-wind-speed-kt --trace-write=/environment/metar/gust-wind-speed-kt --trace-write=/environment/config/boundary/entry/wind-speed-kt --trace-write=/environment/config/boundary/entry/wind-from-heading-deg If you could also describe the specific flight that repros the problem that would be most helpful. Once again, my apologies if this has spoilt anyone's flights. If this is causing people problems, it should probably be backed out. However, I would like to get to the bottom of the issue, as a (working) patch to interpolate METAR is something that we need. -Stuart ___ Yahoo! Mail is the world's favourite email. Don't settle for less, sign up for your free account today http://uk.rd.yahoo.com/evt=44106/*http://uk.docs.yahoo.com/mail/winter07.html - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
Sorry, my FIRST reply was sent unfinished by accident! Here is the right one :-) -- Due to the special weather conditions tonight (only IFR possible) the formerly described effect was very reduced but still visible (!). So the trace (appended) might not be usefull, at least I cannot see anything obvious. I will do another test within the next days with other local METAR values. Therefor my answers to your question are mostly related to former observations: Could you explain what you are seeing in terms of wind direction and strength? Is it: - oscillating in direction and/or speed - rotating continually? - chaotic? Normally oscillating Are you seeing the problem generally, or only when the METAR changes as you fly from one zone to another? Generally, directly after starting from the airport. But - as already mentioned in a former mail - I have to set --enable-real-weather-fetch. Without this there are no problems visible. If it is easily reproducable, Seems depending on the METAR conditions - is there but with different strenght could you run with the following command-line parameters and post the results, either to the list or directly to me: --trace-write=/environment/metar/real-metar --trace-write=/environment/metar/base-wind-range-from --trace-write=/environment/metar/base-wind-range-to --trace-write=/environment/metar/base-wind-speed-kt --trace-write=/environment/metar/gust-wind-speed-kt --trace-write=/environment/config/boundary/entry/wind-speed-kt --trace-write=/environment/config/boundary/entry/wind-from-heading-deg If you could also describe the specific flight that repros the problem that would be most helpful. No specific flight. Mostly from EDDV to EDDW or EDDW to other local airports. I used Bo105, PA 24-250 and PA 28-161. Once again, my apologies if this has spoilt anyone's flights. If this is causing people problems, it should probably be backed out. However, I would like to get to the bottom of the issue, as a (working) patch to interpolate METAR is something that we need. Yes, please try to fix it, NOT remove it. The old classic behaviour (instability of flight when changing METAR zone) was very annoying and unrealistic. -Stuart Thanks anyway for your work in that area :-) Georg EDDW Trace2EDDV.txt.tar.gz Description: application/tgz - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
Laurence Vanek schrieb: Hans Fugal wrote: Flying in the vicinity of KCSQ just now (KCSQ 062055Z AUTO 16006KT 10SM CLR 26/23 A2979 RMK AO2) in both the pa28-161 and j3cub, I noticed some kind of weather-related problems. I'm afraid it was probably weather interpolation-related. My plane was being tossed around like a salad, but as you can see from that METAR there's no mention of severe turbulence. It went away when I chose fair weather scenario. Turning the turbulence down to 0 on the layer I was flying in changed nothing. It wasn't an FPS or machine load problem, which was my first suspect. It didn't feel like real turbulence, but it did feel like fighting walls of weather, as if they were alternating from one to another. FG/OSG (osgviewer) built from CVS from last night (roughly 0800 UTC), which I believe to be current. Happy to compile and try any variant (plib, osg/glut) on request. I can confirm this also. However, I took it as a taste of realism (feature not a bug?). I also had the turb set to zero. As Hans states it seems to conflict with what the METAR is saying. This behaviour occured first here after I compiled OSG CVS with the new weather interpolating patch. AND I have to fetch real weather --enable-real-weather-fetch, what is default for me. NOT using this parameter brings a normal situation back (= smooth flight). This evening I made several testflights from the same airport WITH and WITHOUT real-weather-fetch and it is obvious to me that this bug is related to the latest change (weather interpolating feature). And just adding, when you have this artificial turbulence you can watch your ASI jump between the indicated speeds and when flying pretty low and looking at the electrical windmills you watch them changing their direction very fast. Regards Georg EDDW - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
I did some more investigating, and discovered that it is still happening even fair weather, it's just that the wind speed is low enough that the effect isn't very pronounced. Also, taxi up to a windsock and watch the effect. Incidentally, I was trying to add a layer of clouds with --enable-real-weather-fetch. No matter what weather scenario I had chosen, adding clouds fails to have any effect (and when opening the clouds dialog again, everything is reset to -). I don't know if this is related, but it doesn't seem very desirable to have the weather dialogs do nothing useful when real-weather-fetch is enabled. Semantically, am I right that for weather scenarios, METAR is the real weather, Thunderstorm is thunderstorm-like weather (no relation to real weather?), fair is easy flying (again, no relation to real weather?), and none means no scenario (manual control?). That's what I think they should mean but I'm not convinced that that is what they mean (or anything else that would make sense). On 8/7/07, Georg Vollnhals [EMAIL PROTECTED] wrote: Laurence Vanek schrieb: Hans Fugal wrote: Flying in the vicinity of KCSQ just now (KCSQ 062055Z AUTO 16006KT 10SM CLR 26/23 A2979 RMK AO2) in both the pa28-161 and j3cub, I noticed some kind of weather-related problems. I'm afraid it was probably weather interpolation-related. My plane was being tossed around like a salad, but as you can see from that METAR there's no mention of severe turbulence. It went away when I chose fair weather scenario. Turning the turbulence down to 0 on the layer I was flying in changed nothing. It wasn't an FPS or machine load problem, which was my first suspect. It didn't feel like real turbulence, but it did feel like fighting walls of weather, as if they were alternating from one to another. FG/OSG (osgviewer) built from CVS from last night (roughly 0800 UTC), which I believe to be current. Happy to compile and try any variant (plib, osg/glut) on request. I can confirm this also. However, I took it as a taste of realism (feature not a bug?). I also had the turb set to zero. As Hans states it seems to conflict with what the METAR is saying. This behaviour occured first here after I compiled OSG CVS with the new weather interpolating patch. AND I have to fetch real weather --enable-real-weather-fetch, what is default for me. NOT using this parameter brings a normal situation back (= smooth flight). This evening I made several testflights from the same airport WITH and WITHOUT real-weather-fetch and it is obvious to me that this bug is related to the latest change (weather interpolating feature). And just adding, when you have this artificial turbulence you can watch your ASI jump between the indicated speeds and when flying pretty low and looking at the electrical windmills you watch them changing their direction very fast. Regards Georg EDDW - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- Hans Fugal Fugal Computing - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
[Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
Flying in the vicinity of KCSQ just now (KCSQ 062055Z AUTO 16006KT 10SM CLR 26/23 A2979 RMK AO2) in both the pa28-161 and j3cub, I noticed some kind of weather-related problems. I'm afraid it was probably weather interpolation-related. My plane was being tossed around like a salad, but as you can see from that METAR there's no mention of severe turbulence. It went away when I chose fair weather scenario. Turning the turbulence down to 0 on the layer I was flying in changed nothing. It wasn't an FPS or machine load problem, which was my first suspect. It didn't feel like real turbulence, but it did feel like fighting walls of weather, as if they were alternating from one to another. FG/OSG (osgviewer) built from CVS from last night (roughly 0800 UTC), which I believe to be current. Happy to compile and try any variant (plib, osg/glut) on request. - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Severe Turbulence (Weather Interpolation Problem?)
Hans Fugal wrote: Flying in the vicinity of KCSQ just now (KCSQ 062055Z AUTO 16006KT 10SM CLR 26/23 A2979 RMK AO2) in both the pa28-161 and j3cub, I noticed some kind of weather-related problems. I'm afraid it was probably weather interpolation-related. My plane was being tossed around like a salad, but as you can see from that METAR there's no mention of severe turbulence. It went away when I chose fair weather scenario. Turning the turbulence down to 0 on the layer I was flying in changed nothing. It wasn't an FPS or machine load problem, which was my first suspect. It didn't feel like real turbulence, but it did feel like fighting walls of weather, as if they were alternating from one to another. FG/OSG (osgviewer) built from CVS from last night (roughly 0800 UTC), which I believe to be current. Happy to compile and try any variant (plib, osg/glut) on request. I can confirm this also. However, I took it as a taste of realism (feature not a bug?). I also had the turb set to zero. As Hans states it seems to conflict with what the METAR is saying. - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel