[Flightgear-devel] Fgrun output to fgfs
I would like to make a small mod to Fgrun, I would like explore the options of making it drive fgfs multiple machines rather than just one. I spent some time studying the code today and it looked like it runs fgfs as a child process from the run_posix module. Am I on the right track with this assessment? Possibly the best way might be get it write a file rather then do the rest from a script. -- Regards Harry -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
[Flightgear-devel] Small fgrun 'bug' re Atlas
Hi Fred, As previously stated, thanks for the great WIN32 installer v20110207 ... it works like a charm ;=)) But it seems fgrun adds Atlas - --atlas=socket,out,5,localhost,5505,udp by default, which is not really a problem. But then when you check the Atlas option it adds a 2nd Atlas command - --atlas=socket,out,5,localhost,5500,udp or as you configure it... And when you then uncheck the Atlas option, it removes the 2nd, but not the first Atlas line from the command. How to remove the 'default' Atlas line, if you want no Atlas 'connection'? Regards, Geoff. -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Small fgrun 'bug' re Atlas
Hi Geoff, There is an --atlas option dedicated to Terrasync Regards, -Fred - Geoff McLane a écrit : Hi Fred, As previously stated, thanks for the great WIN32 installer v20110207 ... it works like a charm ;=)) But it seems fgrun adds Atlas - --atlas=socket,out,5,localhost,5505,udp by default, which is not really a problem. But then when you check the Atlas option it adds a 2nd Atlas command - --atlas=socket,out,5,localhost,5500,udp or as you configure it... And when you then uncheck the Atlas option, it removes the 2nd, but not the first Atlas line from the command. How to remove the 'default' Atlas line, if you want no Atlas 'connection'? Regards, Geoff. -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- Frédéric Bouvier http://www.youtube.com/user/fgfred64 Videos -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Sinking feeling - c172 on gravel runway
On Wed, 2011-02-09 at 19:51 +, Martin Spott wrote: I have seen this on various asphalt runways as well... Well, I just experimented by editing apt.dat.gz, and inverted the 'surface' numbers - made the 4204 ft runway 15x a 4, and the short 08x 1902 a 5, and no more tilting ;=)) so far... One might argue that 'Gravel' (5) should be 'harder' than 'Dirt' (4) in the 810 spec, and that 'Turf/grass' (3) should be softer than than either Gravel or Dirt, but perhaps that is a question for Robin Peel. Anyway, this fixed it so I can watch for those pesky kangaroos hopping around ;=)) Regards, Geoff. -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Fgrun output to fgfs
Hi, yes, you're right, fgrun use the code in run_posix.cxx to create a new fgfs process, but only on Linux. You are allowed to specify the command line you want as executable in the first page, so why not starting your script that way ? Regards, -Fred - Harry Campigli harryc...@gmail.com a écrit : I would like to make a small mod to Fgrun, I would like explore the options of making it drive fgfs multiple machines rather than just one. I spent some time studying the code today and it looked like it runs fgfs as a child process from the run_posix module. Am I on the right track with this assessment? Possibly the best way might be get it write a file rather then do the rest from a script. -- Frédéric Bouvier http://www.youtube.com/user/fgfred64 Videos -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Small fgrun 'bug' re Atlas
On Thu, 2011-02-10 at 14:04 +0100, Frederic Bouvier wrote: There is an --atlas option dedicated to Terrasync Hi Fred, Ok, that explains it. Terrasync 'monitors' 5505 for position changes... that make sense. And I can now see unchecking TerraSync:5505 removes that (extra) Atlas option ;=() Looking further into the console window, I can now also see Terrasync is run with -p 5505 Solved, and thanks, Geoff. -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Fgrun output to fgfs
Hey Thank you Frederic, I get the gist of your message, I point fgrun at a script that will accept the argument string, sort out the options then fire of relevant the command to the various machines. i am all for having a custom script to deal with these unorthodox setups than needlessly tinker with code I don't fully understand. Obviously in a multi machine setup I don't want any io and video options passed but the receiving script can be set to filter with that kind of thing. Would I need to fake fgfs back to fgrun with this script? Or just let the script do its thing and exit normally once its kicked off the various machines. Fgrun then just thinks fgfs terminated normally, albeit rather quick? I suspect in a multi machine setup this is best but the script could hang around till the master machine with the sim engine exits. Cheers Harry On Thu, Feb 10, 2011 at 8:09 PM, Frederic Bouvier fredfgf...@free.frwrote: Hi, yes, you're right, fgrun use the code in run_posix.cxx to create a new fgfs process, but only on Linux. You are allowed to specify the command line you want as executable in the first page, so why not starting your script that way ? Regards, -Fred - Harry Campigli harryc...@gmail.com a écrit : I would like to make a small mod to Fgrun, I would like explore the options of making it drive fgfs multiple machines rather than just one. I spent some time studying the code today and it looked like it runs fgfs as a child process from the run_posix module. Am I on the right track with this assessment? Possibly the best way might be get it write a file rather then do the rest from a script. -- Frédéric Bouvier http://www.youtube.com/user/fgfred64 Videos -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- Regards Harry -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Fgrun output to fgfs
It's up to you to block the script or not. fgrun doesn't need anything from the program it launch Regards, -Fred - Harry Campigli a écrit : Hey Thank you Frederic, I get the gist of your message, I point fgrun at a script that will accept the argument string, sort out the options then fire of relevant the command to the various machines. i am all for having a custom script to deal with these unorthodox setups than needlessly tinker with code I don't fully understand. Obviously in a multi machine setup I don't want any io and video options passed but the receiving script can be set to filter with that kind of thing. Would I need to fake fgfs back to fgrun with this script? Or just let the script do its thing and exit normally once its kicked off the various machines. Fgrun then just thinks fgfs terminated normally, albeit rather quick? I suspect in a multi machine setup this is best but the script could hang around till the master machine with the sim engine exits. Cheers Harry On Thu, Feb 10, 2011 at 8:09 PM, Frederic Bouvier fredfgf...@free.fr wrote: Hi, yes, you're right, fgrun use the code in run_posix.cxx to create a new fgfs process, but only on Linux. You are allowed to specify the command line you want as executable in the first page, so why not starting your script that way ? Regards, -Fred - Harry Campigli harryc...@gmail.com a écrit : I would like to make a small mod to Fgrun, I would like explore the options of making it drive fgfs multiple machines rather than just one. I spent some time studying the code today and it looked like it runs fgfs as a child process from the run_posix module. Am I on the right track with this assessment? Possibly the best way might be get it write a file rather then do the rest from a script. -- Frédéric Bouvier http://www.youtube.com/user/fgfred64 Videos -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Sinking feeling - c172 on gravel runway
On Thursday 10 February 2011 06:09:56 Geoff McLane wrote: On Wed, 2011-02-09 at 19:51 +, Martin Spott wrote: I have seen this on various asphalt runways as well... Well, I just experimented by editing apt.dat.gz, and inverted the 'surface' numbers - made the 4204 ft runway 15x a 4, and the short 08x 1902 a 5, and no more tilting ;=)) so far... One might argue that 'Gravel' (5) should be 'harder' than 'Dirt' (4) in the 810 spec, and that 'Turf/grass' (3) should be softer than than either Gravel or Dirt, but perhaps that is a question for Robin Peel. Anyway, this fixed it so I can watch for those pesky kangaroos hopping around ;=)) Regards, Geoff. As far as I know, JSBSim still doesn't know about surfaces? Ron -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Fgrun output to fgfs
Ok Cheers Frederick, A project for tomorrow. Best regards Harry On Thu, Feb 10, 2011 at 9:05 PM, Frederic Bouvier fredfgf...@free.frwrote: It's up to you to block the script or not. fgrun doesn't need anything from the program it launch Regards, -Fred - Harry Campigli a écrit : Hey Thank you Frederic, I get the gist of your message, I point fgrun at a script that will accept the argument string, sort out the options then fire of relevant the command to the various machines. i am all for having a custom script to deal with these unorthodox setups than needlessly tinker with code I don't fully understand. Obviously in a multi machine setup I don't want any io and video options passed but the receiving script can be set to filter with that kind of thing. Would I need to fake fgfs back to fgrun with this script? Or just let the script do its thing and exit normally once its kicked off the various machines. Fgrun then just thinks fgfs terminated normally, albeit rather quick? I suspect in a multi machine setup this is best but the script could hang around till the master machine with the sim engine exits. Cheers Harry On Thu, Feb 10, 2011 at 8:09 PM, Frederic Bouvier fredfgf...@free.frwrote: Hi, yes, you're right, fgrun use the code in run_posix.cxx to create a new fgfs process, but only on Linux. You are allowed to specify the command line you want as executable in the first page, so why not starting your script that way ? Regards, -Fred - Harry Campigli harryc...@gmail.com a écrit : I would like to make a small mod to Fgrun, I would like explore the options of making it drive fgfs multiple machines rather than just one. I spent some time studying the code today and it looked like it runs fgfs as a child process from the run_posix module. Am I on the right track with this assessment? Possibly the best way might be get it write a file rather then do the rest from a script. -- Frédéric Bouvier http://www.youtube.com/user/fgfred64 Videos -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Incorrect conversion used for lbs to gallon of fuel
That was not my point. I was talking about the Nasal error. Nasal runtime error: vector index 7 out of bounds (size: 7) at /wrklvm/FlightGear/FlightGear_CVS/data/Aircraft/Boeing314/Nasal/Boeing314-fuel.nas, line 79 called from: /wrklvm/FlightGear/FlightGear_CVS/data/Aircraft/Boeing314/Nasal/Boeing314-fuel.nas, line 70 called from: /wrklvm/FlightGear/FlightGear_CVS/data/Aircraft/Boeing314/Nasal/Boeing314-fuel.nas, line 59 called from: /wrklvm/FlightGear/FlightGear_CVS/data/Aircraft/Boeing314/Nasal/Boeing314-fuel.nas, line 19 called from: /wrklvm/FlightGear/FlightGear_CVS/data/Aircraft/Boeing314/Nasal/Boeing314.nas, line 92 called from: /wrklvm/FlightGear/FlightGear_CVS/data/Aircraft/Boeing314/Nasal/Boeing314.nas, line 111 called from: /wrklvm/FlightGear/FlightGear_CVS/data/Aircraft/Boeing314/Nasal/Boeing314.nas, line 17 called from: /wrklvm/FlightGear/FlightGear_CVS/data/Aircraft/Boeing314/Nasal/Boeing314.nas, line 125 called from: /wrklvm/FlightGear/FlightGear_CVS/data/Nasal/globals.nas, line 100 2011/2/8 Csaba Halász csaba.hal...@gmail.com On Tue, Feb 8, 2011 at 10:05 PM, henri orange hohora...@gmail.com wrote: Will it solved the Boeing314 issue ? The Boeing314 in GIT doesn't even get that far, it is missing a file Nasal/Boeing314-limits.xml Commenting the reference out, the aircraft at least takes off. -- Csaba/Jester -- Best regards, Henri, aka Alva Official grtux hangar maintainer -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Incorrect conversion used for lbs to gallon of fuel
On Thu, Feb 10, 2011 at 3:21 PM, henri orange hohora...@gmail.com wrote: That was not my point. I was talking about the Nasal error. I know. I just pointed out that the GIT version is broken anyway. BUT working around the missing file, it DOES now fly, without the fuel nasal error. -- Csaba/Jester -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Sinking feeling - c172 on gravel runway
Correct. JSBSim itself makes no distinction between ground materials (hence the reason why some aircrafts are able to land on water). This can however be managed with Nasal scripts. So I would say that this issue is likely located in one of the C172 Nasal scripts. Bertrand Le 10 févr. 2011 09:19, Ron Jensen w...@jentronics.com a écrit : On Thursday 10 February 2011 06:09:56 Geoff McLane wrote: On Wed, 2011-02-09 at 19:51 +, Marti... As far as I know, JSBSim still doesn't know about surfaces? Ron -- The ultimate all-in-... -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Sinking feeling - c172 on gravel runway
By the way we can play with the gear Z position (gear/unit/z-position) and the well known terrain nasal script (geodinfo) which expose the terrain load-resistance. Though, that feature is only eye candy. More constructive, to use the specifics friction_factor, rolling_friction experimented with the grtux aircrafts. 2011/2/10 Bertrand Coconnier bcoco...@gmail.com Correct. JSBSim itself makes no distinction between ground materials (hence the reason why some aircrafts are able to land on water). This can however be managed with Nasal scripts. So I would say that this issue is likely located in one of the C172 Nasal scripts. Bertrand Le 10 févr. 2011 09:19, Ron Jensen w...@jentronics.com a écrit : On Thursday 10 February 2011 06:09:56 Geoff McLane wrote: On Wed, 2011-02-09 at 19:51 +, Marti... As far as I know, JSBSim still doesn't know about surfaces? Ron -- The ultimate all-in-... -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- Best regards, Henri, aka Alva Official grtux hangar maintainer -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Sinking feeling - c172 on gravel runway
Hi, RE: http://www.geoffair.net/tmp/tilted-001.png Thank you. All good points... Am presently Terrasync e000n40 so I can try the Grass (3) at LFPZ Saint Cyr l Ecole, where I am sure I have seen this tilting back several times before... Also, it does not happen EVERY time, in each of 3 machines, 3 OSes... but when I get the situation, it seems very repeatable... It happens after the begin fade of the splash screen, and the first view out of the a/c always seems flat and level... If it is going to happen, it is after a second or so, then slowly over the next 4-5 seconds... up to about 5 degrees, but can be less... Sorry, all just guesses at this stage... and have yet to try other a/c... Thanks again, and regards, Geoff. -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Near Clipping in the virtual cockpit: please help!!
Hi Paul, Here is an interesting link that explains the core of the problem. http://www.sjbaker.org/steve/omniv/love_your_z_buffer.html It's been a while, but one technique for dealing with this is to clear the zbuffer and reset the near plane location when drawing different portions of the scene. This can work well if you can divide up your scene into stuff that is further away (like terrain) and stuff that is closer (like the cockpit) and then carefully pick the order you draw the different portions of your scene graph. Regards, Curt. On Thu, Feb 10, 2011 at 1:04 AM, Paul Guhl ad...@simtechnologies.de wrote: Hello! I have unsolved problem driving me nuts for weeks now. In my opengl application the terrain presentation is flickering very frequently (either as whole our 1/3 of all polygons). The problem disappears as soon as i increase the near clipping value. The problem root is the z buffer precision, but greater near clipping in turn truncates much of the virtual cockpit display. How did you solve the clipping problem in flight gear??? Thanks in advance Paul -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- Curtis Olson: http://www.atiak.com - http://aem.umn.edu/~uav/ http://www.flightgear.org - http://www.flightgear.org/blogs/category/curt/http://www.flightgear.org/blogs/category/personal/curt/ -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Sinking feeling - c172 on gravel runway
On 02/10/2011 08:15 AM, Geoff McLane wrote: Also, it does not happen EVERY time, in each of 3 machines, 3 OSes... but when I get the situation, it seems very repeatable... Is there any chance this depends on having a _tailwind_ or something like that? Crosswinds and tailwinds can cause some mighty peculiar behavior: http://code.google.com/p/flightgear-bugs/issues/detail?id=250 -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Sinking feeling - c172 on gravel runway
On Thu, 10 Feb 2011 09:27:55 -0700, John wrote in message 4d54120b.1090...@av8n.com: On 02/10/2011 08:15 AM, Geoff McLane wrote: Also, it does not happen EVERY time, in each of 3 machines, 3 OSes... but when I get the situation, it seems very repeatable... Is there any chance this depends on having a _tailwind_ or something like that? Crosswinds and tailwinds can cause some mighty peculiar behavior: http://code.google.com/p/flightgear-bugs/issues/detail?id=250 ..I cannot reproduce this: arnt@celsius:~/FG-git$ ./run_fgfs.sh --geometry=1920x1200 --httpd= \ --jpg-httpd=9876 --prop:controls/gear/brake-parking=1 \ --prop:sim/frame-rate=true --fov=90 --enable-fullscreen \ --timeofday=noon --aircraft=c172p --airport=KSFO --runway=19L \ --disable-real-weather-fetch --metar= 012345Z 07315KT 50SM \ FEW015 SCT025 BKN035 OVC090 15/M01 A3092 [1] 3843 arnt@celsius:~/FG-git$ Processing command line arguments Fatal error: Failed to open file at 012345Z (received from SimGear XML Parser) [1]+ Done./run_fgfs.sh --geometry=1920x1200 --httpd= --jpg-httpd=9876 --prop:controls/gear/brake-parking=1 --prop:sim/frame-rate=true --fov=90 --enable-fullscreen --timeofday=noon --aircraft=c172p --airport=KSFO --runway=19L --disable-real-weather-fetch --metar= 012345Z 07315KT 50SM FEW015 SCT025 BKN035 OVC090 15/M01 A3092 arnt@celsius:~/FG-git$ -- ..med vennlig hilsen = with Kind Regards from Arnt Karlsen ...with a number of polar bear hunters in his ancestry... Scenarios always come in sets of three: best case, worst case, and just in case. -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Sinking feeling - c172 on gravel runway
Hi, Re: http://www.geoffair.net/tmp/tilted-001.png Again, thanks for the multiple inputs... 1. __tailwind__ (a) John, I wanted to try your metar to see if it looked related, but like Arnt I can not get the metar string through the command parser! Always - Fatal error: Failed to open file at 012345Z (received from SimGear XML Parser) What are we doing wrong? I tried, like you and Arnt put - --metar= 012345Z etc then --metar= 012345Z etc even '--metar= 012345Z etc' --metar=' 012345Z etc' Always the same error message and fgfs aborts... (b) But also, I think in most of the trials I had disabled real weather, or at least not enabled it, so do not think my case is related to 'wind', but maybe... 2. LFPZ (Grass) On all 4 runways, all looked ok, although sometimes there was some unusual a/c 'rocking', after the splash cleared, usually first down, then up, which 'felt' like what happens at YGIL, but it seemed to stay on 3 feet ;=)) But that was in an old XP machine, where I do not see a frame rate much above about 4 or 5, which makes it virtually unplayable, so am presently redoing a full git update, and SG/FG rebuild in Ubuntu, to try there... From what Henri said, it certainly seems possible that nasal is dickering with the gear Z position (gear/unit/z-position) ..., or something, so more to explore here... I can see in c172p/Nasal/action_sim.nas that it, in the update_actions function, gets lots of gear related nodes like - var dhN_ft = props.globals.getNode(gear/gear[0]/compression-ft, 1); var dhR_ft = props.globals.getNode(gear/gear[2]/compression-ft, 1); var dhL_ft = props.globals.getNode(gear/gear[1]/compression-ft, 1); var propGear0 = props.globals.getNode(gear/gear[0], 1); var propGear1 = props.globals.getNode(gear/gear[1], 1); var propGear2 = props.globals.getNode(gear/gear[2], 1); var nose_link_rot = propGear0.getNode(compression-rotation-deg, 1); var left_main_rot = propGear1.getNode(compression-rotation-deg, 1); var right_main_rot = propGear2.getNode(compression-rotation-deg, 1); and later sets the latter 3 to calculated values... At the end sets itself up to repeat on each frame - settimer(update_actions, 0); And ties itself into - setlistener(sim/signals/fdm-initialized, init_actions); and setlistener(sim/signals/reset, init_actions); H... can not yet see where the r/w 'surface' is in this, but some things to try here, to at least rule out this particular nasal, like comment out 1 or all of - nose_link_rot.setDoubleValue(theta); right_main_rot.setDoubleValue(right_alpha_deg); left_main_rot.setDoubleValue(left_alpha_deg); But out of time today, but will be back tomorrow, after eating and sleeping on it... I hope... ;=)) Thanks for all the input, and any other ideas welcome... I look forward to a quiet life at YGIL ;=)) Usually do not have many visitors, except the kangas... Regards, Geoff. -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Sinking feeling - c172 on gravel runway
Geoff McLane wrote: 1. __tailwind__ I would assume that if you don't specify the runway on startup, you'll automagically get placed against the wind, as you would do in real life. If I'm correct, then tailwind does not apply. Cheers, martin. -- Unix _IS_ user friendly - it's just selective about who its friends are ! -- -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN
Hi Henri, I think I found the error, it was in JSBSim's FGForce class. I've proposed a fix on the JSBSim-devel mailing list. Best regards, Andreas -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
[Flightgear-devel] Sgtext
Hi guys, I.ve been working at changing over the Citation X EFIS to use 3d text ,but having a few problems. The property scale and offset don't work here , and formating doesnt behave like I expected ... for example , format%2.0f/format for altitude at 3029.23341234 prints 3029 . Scaling the property by 0.01 prints the same. Are there any working examples I can compare with ? I'm also wondering if it's an OSG error, (2.9.10), but I'm not to familiar with Callbacks the text code has several if(numeric) checks , but I can't figure out where numeric is set . Thanks in advance if anyone can shed light on this. Syd -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
[Flightgear-devel] SimGear bug? read-only properties tied to object methods not initialized
Hi all - here comes a tricky one: (good bugs are usually tricky - not everything tricky is necessarily a bug) Imagine an arbitrary PropertyList config file like this: PropertyList foo type=double3.14/foo /PropertyList and a code fragment like class MyClass { public: double getter() const { return foo; } void setter( double d ) { foo = d; } void dosomethingfancy(); private: double foo; }; void MyClass::dosomethingfancy() { someBaseNode-tie( foo, this, MyClass:getter, MyClass::setter ); } Ignore the fact, that it doesn't make much sense here - it's just an example. One would expect that foo in MyClass (and the value of the thereafter tied property) gets initialized to the value of foo from the config file which is 3.14. This actually works fine. Now, modify the xml file to look like this: PropertyList foo type=double write=n3.14/foo /PropertyList (note: foo is now read-only) The value of the foo-property now remains at an unchangeable value of zero, it no longer gets initialized to the provided value in foo. That is because the property is declared read-only and the setter method never gets called. In simgear/props/props.hxx in SGPropertyNode::tie() some code exist to set the value of a property after being tied, to the default value using setValue(). However, setValue() is smart enough to check for a read-only property and refuses to change a property value in that case. A fix might be to temporarily set the WRITE-flag for the property in the tie() method like this: (fragment from props.hxx, SGPropertyNode::tie(const SGRawValueT rawValue, bool useDefault), line 1789ff if (useDefault) { int save_attributes = getAttributes(); // save current write-flag setAttribute( WRITE, true ); // write-enable the property setValue(old_val); // set the old value setAttributes( save_attributes ); // restore write-flag } instead of if (useDefault) { setValue(old_val); // noop for read-only property } Comments? Torsten -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Sgtext
Hi guys, I.ve been working at changing over the Citation X EFIS to use 3d text ,but having a few problems. The property scale and offset don't work here , and formating doesnt behave like I expected ... for example , format%2.0f/format for altitude at 3029.23341234 prints 3029 . Scaling the property by 0.01 prints the same. Are there any working examples I can compare with ? I'm also wondering if it's an OSG error, (2.9.10), but I'm not to familiar with Callbacks the text code has several if(numeric) checks , but I can't figure out where numeric is set . Thanks in advance if anyone can shed light on this. Syd Hi Syd, Anders has an excellent example in his ZLT-NT. Check folder Aircraft/ZLT- NT/Models/Instruments/CenterMFD and look into screen-engines.xml and display- text.xml. The numeric property in the UpdateCallback is set to true if the type element is set to number-value. (Line 218 in SGText.cxx). Hope this helps Torsten -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Sgtext
Hi Syd, I use the text (aka OSG Text) a lot in Aircraft/A380/XML/Flightdeck/Pedestal/mcdu_new_display_01.xml But if you are after a zero filled format, have a look in Aircraft/A380/XML/Glareshield/fcu.xml text nameFCU.hdg.degs/name axis-alignmentyz-plane/axis-alignment alignmentright-bottom/alignment offsets pitch-deg-90.0/pitch-deg x-m-0.068/x-m y-m-0.035/y-m z-m-0.0009/z-m /offsets typenumber-value/type property type=string/instrumentation/afs/heading-bug-deg/property format type=string%03.0f/format fontled.txf/font character-size type=double0.009/character-size /text I'm not sure what you were after with %2.0f but the example you give looks correct to me. I don't think I've had to use scaling on any text values though. S. On Thu, 2011-02-10 at 13:26 -0700, syd adams wrote: Hi guys, I.ve been working at changing over the Citation X EFIS to use 3d text ,but having a few problems. The property scale and offset don't work here , and formating doesnt behave like I expected ... for example , format%2.0f/format for altitude at 3029.23341234 prints 3029 . Scaling the property by 0.01 prints the same. Are there any working examples I can compare with ? I'm also wondering if it's an OSG error, (2.9.10), but I'm not to familiar with Callbacks the text code has several if(numeric) checks , but I can't figure out where numeric is set . Thanks in advance if anyone can shed light on this. Syd -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
[Flightgear-devel] Blender question?
I have a hopefully quick question. I've generated a 3d model mesh in ac3d format. I'm doing this from a perl script and I posted some pictures and details of the actual model here: http://www.flightgear.org/blogs/curt/uas/misc/3d-modelling-with-perl/ My script just generates the left half of the model. I assumed I could just import this into blender, duplicate the half and mirror it and produce the whole model. I'm new to blender, but I managed to duplicate the side and mirror it and the mesh looks perfect. My problem is that when I export the full model, the mirrored half is black from the outside. When I look inside of it, it's shaded properly. It appears that when I mirrored the surface, the face ordering didn't change so the mirrored half is inside out. I've been trying every possible face/normal/edge option I can find in blender and haven't been able to figure out how to get my faces back the right way. The original half of course looks just fine. It's probably something super simple, but I've googled and haven't found the right set of keywords I guess. Is there an easy way to get all my faces the right way so both sides of my model are right side out and look correct? Thanks, Curt. -- Curtis Olson: http://www.atiak.com - http://aem.umn.edu/~uav/ http://www.flightgear.org - http://www.flightgear.org/blogs/category/curt/http://www.flightgear.org/blogs/category/personal/curt/ -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Blender question?
On Thursday, February 10, 2011 01:24:32 PM Curtis Olson wrote: I have a hopefully quick question. I've generated a 3d model mesh in ac3d format. I'm doing this from a perl script and I posted some pictures and details of the actual model here: http://www.flightgear.org/blogs/curt/uas/misc/3d-modelling-with-perl/ My script just generates the left half of the model. I assumed I could just import this into blender, duplicate the half and mirror it and produce the whole model. I'm new to blender, but I managed to duplicate the side and mirror it and the mesh looks perfect. My problem is that when I export the full model, the mirrored half is black from the outside. When I look inside of it, it's shaded properly. It appears that when I mirrored the surface, the face ordering didn't change so the mirrored half is inside out. I've been trying every possible face/normal/edge option I can find in blender and haven't been able to figure out how to get my faces back the right way. The original half of course looks just fine. It's probably something super simple, but I've googled and haven't found the right set of keywords I guess. Is there an easy way to get all my faces the right way so both sides of my model are right side out and look correct? Thanks, Curt. First you need to make sure that the surfaces are single sided. At that point the part that is black from the outside should become transparant from the outside. Then you need to flip the normals for those faces. You might be able to do this by using normals recalculate outside in edit mode. When you are in Edit mode in the 3D part of the display you should be able to activate the buttons menu in the other part of the screen and it will allow you to turn on displaying the normals vectors in the 3D screen. This will let you know for sure which way the normals are pointing. Hal -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Blender question?
On Thu, 10 Feb 2011, Curtis Olson wrote: It's probably something super simple, but I've googled and haven't found the right set of keywords I guess. Is there an easy way to get all my faces the right way so both sides of my model are right side out and look correct? I'm no blender expert but what I do after mirroring is selecting the faces of the mirrored part and flip the normals ('w'-flip normals). There used to be some apply transformation command (alt+a?) but I have not managed to find it in recent times. OTOH for your model you could also add the mirror modifier to the orginal half and leave it to the AC3D exporter to actually mirror the mesh - that usually works without problems. Cheers, Anders -- --- Anders Gidenstam WWW: http://www.gidenstam.org/FlightGear/ -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN
Hi, Andreas Thanks a lot. However , with the Catalina last updated version: https://sites.google.com/site/grtuxhangar/home/download/PBY-Catalina.tar.bz2?attredirects=0d=1 On my system, the error vanished, i don't understand why. Anyhow your jsbsim fix should solve , or at least reduce the bad luck to get again such error. I 'll give it a try . 2011/2/10 Andreas Gaeb a.g...@web.de Hi Henri, I think I found the error, it was in JSBSim's FGForce class. I've proposed a fix on the JSBSim-devel mailing list. Best regards, Andreas -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- Best regards, Henri, aka Alva Official grtux hangar maintainer -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Blender question?
Hey Curt, This is the process I use in Blender 2.49 - its slightly different but similar for Blender 2.5x Use face select mode, select the faces you need to flip, and in 'mesh tools' click 'flip normals', then re-save your model - that *should* do the trick. Another useful trick is to click 'show normals' - you'll then see a small blue line extending from the centre of each face in the positive direction - if the normals point inward then they need to be flipped. The normal size can be changed if they are too hard to see. Regards, Chris Wilkinson, YBBN/BNE. From: Curtis Olson curtol...@gmail.com To: FlightGear developers discussions flightgear-devel@lists.sourceforge.net Sent: Fri, 11 February, 2011 7:24:32 AM Subject: [Flightgear-devel] Blender question? I have a hopefully quick question. I've generated a 3d model mesh in ac3d format. I'm doing this from a perl script and I posted some pictures and details of the actual model here: http://www.flightgear.org/blogs/curt/uas/misc/3d-modelling-with-perl/ My script just generates the left half of the model. I assumed I could just import this into blender, duplicate the half and mirror it and produce the whole model. I'm new to blender, but I managed to duplicate the side and mirror it and the mesh looks perfect. My problem is that when I export the full model, the mirrored half is black from the outside. When I look inside of it, it's shaded properly. It appears that when I mirrored the surface, the face ordering didn't change so the mirrored half is inside out. I've been trying every possible face/normal/edge option I can find in blender and haven't been able to figure out how to get my faces back the right way. The original half of course looks just fine. It's probably something super simple, but I've googled and haven't found the right set of keywords I guess. Is there an easy way to get all my faces the right way so both sides of my model are right side out and look correct? Thanks, Curt. -- Curtis Olson: http://www.atiak.com - http://aem.umn.edu/~uav/ http://www.flightgear.org - http://www.flightgear.org/blogs/category/curt/ -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Blender question?
Hi Chris, I turned on normals and they are all pointing outwards on both halves of my model. It looks good in blender, but when I export it to 3ds and open it up in osgviewer, the mirrored half is still inside out. Curt. On Thu, Feb 10, 2011 at 3:45 PM, Chris Wilkinson wrote: Hey Curt, This is the process I use in Blender 2.49 - its slightly different but similar for Blender 2.5x Use face select mode, select the faces you need to flip, and in 'mesh tools' click 'flip normals', then re-save your model - that *should* do the trick. Another useful trick is to click 'show normals' - you'll then see a small blue line extending from the centre of each face in the positive direction - if the normals point inward then they need to be flipped. The normal size can be changed if they are too hard to see. Regards, Chris Wilkinson, YBBN/BNE. -- *From:* Curtis Olson curtol...@gmail.com *To:* FlightGear developers discussions flightgear-devel@lists.sourceforge.net *Sent:* Fri, 11 February, 2011 7:24:32 AM *Subject:* [Flightgear-devel] Blender question? I have a hopefully quick question. I've generated a 3d model mesh in ac3d format. I'm doing this from a perl script and I posted some pictures and details of the actual model here: http://www.flightgear.org/blogs/curt/uas/misc/3d-modelling-with-perl/ My script just generates the left half of the model. I assumed I could just import this into blender, duplicate the half and mirror it and produce the whole model. I'm new to blender, but I managed to duplicate the side and mirror it and the mesh looks perfect. My problem is that when I export the full model, the mirrored half is black from the outside. When I look inside of it, it's shaded properly. It appears that when I mirrored the surface, the face ordering didn't change so the mirrored half is inside out. I've been trying every possible face/normal/edge option I can find in blender and haven't been able to figure out how to get my faces back the right way. The original half of course looks just fine. It's probably something super simple, but I've googled and haven't found the right set of keywords I guess. Is there an easy way to get all my faces the right way so both sides of my model are right side out and look correct? Thanks, Curt. -- Curtis Olson: http://www.atiak.com - http://aem.umn.edu/~uav/ http://www.flightgear.org - http://www.flightgear.org/blogs/category/curt/http://www.flightgear.org/blogs/category/personal/curt/ -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- Curtis Olson: http://www.atiak.com - http://aem.umn.edu/~uav/ http://www.flightgear.org - http://www.flightgear.org/blogs/category/curt/http://www.flightgear.org/blogs/category/personal/curt/ -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Bo105: Nasal runtime error: props.setDoubleValue() passed a NaN
Stay tuned, Quick update: I found the reason for this bug (see previous post SimGear bug? read-only properties tied to object methods not initialized) so there is a good chance for a fix soon. Torsten -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Blender question?
Curt, To add to what Anders and Chris have said: Mirroring an object reverses the order of vertices in faces, which results in polygon normals pointing in the opposite direction-- outside becomes inside. If you're mirroring in Edit mode, you should only need to do the flip normals command. I.e., with the mirrored elements selected, hit the 'w' key and select 'Flip Normals'. After doing this, consider selecting everything and doing a w-Remove Doubles. This will take your two halves and create a single unified whole, eliminating overlaping edges and vertices, which helps avoid unwanted seams when rendered. Anders mentions the ctrl-a command. When mirroring an object in Object mode to create a second object for later merging with the first, it's a good idea to select the mirrored object and hit ctrl-a-Scale and Rotation to ObData. Blender keeps a stack of modifying operators on an object, but isn't perfectly consistent about considering stacked operators. This can yield unexpected results when mirroring and flipping normals. So I recommend using the ctrl-a command at the object level to apply and clear operations to the object, then go into Edit mode and do the flip-normals thing. It can save you some headaches, particularly if an object seems stubborn about being inside-out when exported. -Gary, aka Buckaroo On Thu, Feb 10, 2011 at 4:55 PM, Curtis Olson curtol...@gmail.com wrote: Hi Chris, I turned on normals and they are all pointing outwards on both halves of my model. It looks good in blender, but when I export it to 3ds and open it up in osgviewer, the mirrored half is still inside out. Curt. On Thu, Feb 10, 2011 at 3:45 PM, Chris Wilkinson wrote: Hey Curt, This is the process I use in Blender 2.49 - its slightly different but similar for Blender 2.5x Use face select mode, select the faces you need to flip, and in 'mesh tools' click 'flip normals', then re-save your model - that *should* do the trick. Another useful trick is to click 'show normals' - you'll then see a small blue line extending from the centre of each face in the positive direction - if the normals point inward then they need to be flipped. The normal size can be changed if they are too hard to see. Regards, Chris Wilkinson, YBBN/BNE. From: Curtis Olson curtol...@gmail.com To: FlightGear developers discussions flightgear-devel@lists.sourceforge.net Sent: Fri, 11 February, 2011 7:24:32 AM Subject: [Flightgear-devel] Blender question? I have a hopefully quick question. I've generated a 3d model mesh in ac3d format. I'm doing this from a perl script and I posted some pictures and details of the actual model here: http://www.flightgear.org/blogs/curt/uas/misc/3d-modelling-with-perl/ My script just generates the left half of the model. I assumed I could just import this into blender, duplicate the half and mirror it and produce the whole model. I'm new to blender, but I managed to duplicate the side and mirror it and the mesh looks perfect. My problem is that when I export the full model, the mirrored half is black from the outside. When I look inside of it, it's shaded properly. It appears that when I mirrored the surface, the face ordering didn't change so the mirrored half is inside out. I've been trying every possible face/normal/edge option I can find in blender and haven't been able to figure out how to get my faces back the right way. The original half of course looks just fine. It's probably something super simple, but I've googled and haven't found the right set of keywords I guess. Is there an easy way to get all my faces the right way so both sides of my model are right side out and look correct? Thanks, Curt. -- Curtis Olson: http://www.atiak.com - http://aem.umn.edu/~uav/ http://www.flightgear.org - http://www.flightgear.org/blogs/category/curt/ -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- Curtis Olson: http://www.atiak.com - http://aem.umn.edu/~uav/ http://www.flightgear.org - http://www.flightgear.org/blogs/category/curt/ -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance.
Re: [Flightgear-devel] Sgtext
Thanks guys I'll look through those examples... The reason I need to split the altitude is the PFD display scrolls less than 100 values in steps of 10 and in smaller text. 100 and greater digits are larger non scrolling text ... easy to do with textures and textranslate, but I'm attempting a 3d text only version if possible. The zero and space padding works here , and precision below the decimal point. I still suspect I,ve got an error further up the xml file , just can't see it. On Thu, Feb 10, 2011 at 2:10 PM, Scott Hamilton scott.hamil...@popplanet.biz wrote: Hi Syd, I use the text (aka OSG Text) a lot in Aircraft/A380/XML/Flightdeck/Pedestal/mcdu_new_display_01.xml But if you are after a zero filled format, have a look in Aircraft/A380/XML/Glareshield/fcu.xml text nameFCU.hdg.degs/name axis-alignmentyz-plane/axis-alignment alignmentright-bottom/alignment offsets pitch-deg-90.0/pitch-deg x-m-0.068/x-m y-m-0.035/y-m z-m-0.0009/z-m /offsets typenumber-value/type property type=string/instrumentation/afs/heading-bug-deg/property format type=string%03.0f/format fontled.txf/font character-size type=double0.009/character-size /text I'm not sure what you were after with %2.0f but the example you give looks correct to me. I don't think I've had to use scaling on any text values though. S. On Thu, 2011-02-10 at 13:26 -0700, syd adams wrote: Hi guys, I.ve been working at changing over the Citation X EFIS to use 3d text ,but having a few problems. The property scale and offset don't work here , and formating doesnt behave like I expected ... for example , format%2.0f/format for altitude at 3029.23341234 prints 3029 . Scaling the property by 0.01 prints the same. Are there any working examples I can compare with ? I'm also wondering if it's an OSG error, (2.9.10), but I'm not to familiar with Callbacks the text code has several if(numeric) checks , but I can't figure out where numeric is set . Thanks in advance if anyone can shed light on this. Syd -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel