[Flightgear-devel] AI FlightPlan execution time
Hey there, In the last couple of months, I've been noticing a bit of an irregularity in frame rates, as the AIModel processing load goes up (into the 400+ nr of models). While investigating the New Zealand lock-up bug (see yesterday's post), I found that flightplan processing is scheduled to be executed every 100 ms. Because this execution is triggered from one while loop in the AIManager, flightplan execution has a tendency to synchronize, which I believe is a major contributer to the observed irregularities. For this reason, I'm working on adding a little bit of randomization into to the execution time, much similar to how the ground elevation code for AIModels is scheduled by a random number. Even if I knock the schedule off by just a few random ms, each time, the statistical properties of the random number generator will cause the flightplan processing time of each AIAircraft to be smoothly distributed across multiple executions of the AIManager::update() function. At this point, I don't see any adverse effects of such randomization, but if there are, please let me know. Cheers, Durk - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] ..does SimGear(_plib) build at all --with-jpeg-factory, was: Issues compiling taxidraw...
Hi, Arnt Karlsen wrote: ..and with the @ in place on line 22 in trunk, works ok for all and without the @ something does not work? So after all it was the spaces and the quotes? Enabled products: Atlas CGAL FLU FlightGear FlightGear_base OpenProducer OpenSceneGraph OpenThreads SimGear TaxiDraw TerraGear fgrun fgsd gpc gts jsbsim newmat plib shapelib Disabled products: FlightGear_plib SimGear_plib alut openal [EMAIL PROTECTED]:/opt/src/fgfsbuilder/trunk$ Enabled products: FlightGear_base FlightGear_plib SimGear_plib plib Disabled products: Atlas CGAL FLU FlightGear OpenProducer OpenSceneGraph OpenThreads SimGear TaxiDraw TerraGear alut fgrun fgsd gpc gts newmat openal shapelib [EMAIL PROTECTED]:/opt/src/fgfsbuilder/stable$ ..why _is_ alut and openal there? ;o) Because fgfs-builder is not Debian-specific. If it were Debian-specific, I could also remove gts and some other libs from the list... Cheers, Ralf - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
[Flightgear-devel] RFC: Toward a new policy for AIAircaft commits
Hi, As most developers will know, the AIModels subsystem is used as the main animation system for external traffic in FlightGear. Currently, it subserves three functions: 1) Direct animations through scripts, 2) Animation of Multiplayer aircraft, 3) Animations for, and interactions with a semi-intelligently operating traffic system. Maik Justus has recently contributed a patch that allows the AIModels system to load a dedicated AI version of the Flyable aircraft, provided this version is present in the base package. In addition, I currently have a patch nearly ready that will allow FlightGear to scan the data/AI/Aircraft directory for xml files containing traffic patterns. Finally, there is some progress on a program that will be able to read airline timetable-like databases and convert these into AI traffic files. All 'n' all, these developments will allow us to have a lot more flexibility in the AI system, and will open up the possibility to actually get some substantial amounts of traffic into FlightGear, other than the small demos I built a few years ago. For this reason, I believe it's getting time to start adding (some) AI aircraft to the base repository. Once I have the few remaining rough edged of the new traffic scan patch ironed out, my plan would be to begin by adding the AI 737 to the base package, along with a few liveries for which I have traffic available (United, Malev, and Lufhansa). Because adding AI aircraft to the base package would mean adding quite a bit of binary texture files, I'd like keep the number of revisions for these files down to a minimum, so my proposal is to commit only those textures that are reasonably finished. I'd be interested to hear other people's thoughts and opinions before proceeding. Cheers, Durk - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
[Flightgear-devel] Threadsafe reference counts
Hi, I have now committed a patch to our reference counter that will make reference counting thread safe. As far as I could test or have testers, this patch uses gcc's or MSVC's atomic compiler intrinsics for that task. IMO that is actually the least intrusive way to ensure thread safety. I hope that these compilers also know much about that builtin functions to do some optimizations with them. For *all* other architectures this falls back to using mutexes which is quite heavyweight for that task. That means it will work, but probably not really fast. I believe that I have covered the most important architectures/compilers with win32 and gcc i386/x86_64. But the next important ones from my point of view are the ppc/macos X and the IRIX versions. I believe that these also have some intrinsics or assembly instructions that guarantee atomic changes to an (unsigned) int value, but I have no macos machine present. So If anybody knows how to do that for his favourite architecture, feel free to provide patches/check that in ... Greetings Mathias - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
[Flightgear-devel] PRE_OSG cvs BUG in VSI with code fix
Hi all, Being really new to this (6 days), after building my own version, I found all/no aircraft didn't have a working vsi gauge. After a lot of hair pulling, beer, and more beer, I found why. The pivotal point was seeing I had two 'system/static' branches - one capitised. There is a typo in: src/Instrumentation/vertical_speed_indicator.cxx line 17: _static_pressure(node-getStringValue(static-pressure, /Systems/static/pressure-inhg)) should be of course: _static_pressure(node-getStringValue(static-pressure, /systems/static/pressure-inhg)) Regards, Nick - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
[Flightgear-devel] 787 Revision
I have made some revisions to the 787 model. The flaps were modified. http://www.boeing.com/commercial/787family/news/2006/q4/061206c_pr.htmlhttp://www.boeing.com/commercial/787family/news/2006/q4/061206c_pr.html The exhaust was removed, doube vertices from the fuselage and bodyfairing were merged, and instruments, strobe lights and beacon lights were fixed. Also, view distances were adjusted. Josh - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] 787 Revision
I forgot to add that the link is the same. http://www.golffoxtrotsierra.741.com/787.ziphttp://www.golffoxtrotsierra741.com/787.zip Josh- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] 787 Revision
On Wednesday 27 December 2006 16:14, JOSHUA WILSON wrote: I have made some revisions to the 787 model. The flaps were modified. exhaust was removed, doube vertices from the fuselage and bodyfairing were merged, and instruments, strobe lights and beacon lights were fixed. Also, view distances were adjusted. Hi Josh, any reason for the exhaust effect being removed other than the fact it doesn't yet work with the OSG version of FG? I thought it looked really good, and I'm sure that animation will be reinstated in FG-OSG in due course so don't remove it on that account. If it was removed from personal preference... well, you're the author of the model, what you say, goes :-) Cheers, AJ - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] 787 Revision
AJ: Thank you for your input. I have returned the exhaust feature to the model. Josh- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] ..does SimGear(_plib) build at all --with-jpeg-factory, was: Issues compiling taxidraw...
On Wed, 27 Dec 2006 09:53:18 +0100, Ralf wrote in message [EMAIL PROTECTED]: Hi, Arnt Karlsen wrote: ..and with the @ in place on line 22 in trunk, works ok for all and without the @ something does not work? ..does not appear to matter in stable, and I forgot to chatge it in trunk and it still worked nicely. So after all it was the spaces and the quotes? ..yup, looks that way, spaces outside the arguments inside the quotes. I have a space between --with-jpeg-factory and the OSG arguments, was read correctly this time, AFAICT. Enabled products: Atlas CGAL FLU FlightGear FlightGear_base OpenProducer OpenSceneGraph OpenThreads SimGear TaxiDraw TerraGear fgrun fgsd gpc gts jsbsim newmat plib shapelib Disabled products: FlightGear_plib SimGear_plib alut openal [EMAIL PROTECTED]:/opt/src/fgfsbuilder/trunk$ Enabled products: FlightGear_base FlightGear_plib SimGear_plib plib Disabled products: Atlas CGAL FLU FlightGear OpenProducer OpenSceneGraph OpenThreads SimGear TaxiDraw TerraGear alut fgrun fgsd gpc gts newmat openal shapelib [EMAIL PROTECTED]:/opt/src/fgfsbuilder/stable$ ..why _is_ alut and openal there? ;o) Because fgfs-builder is not Debian-specific. If it were Debian-specific, I could also remove gts and some other libs from the list... ..mmm. make all-debs ? ;o) -- ..med vennlig hilsen = with Kind Regards from Arnt... ;o) ...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. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel