FLIGHT GEAR "This truly is as real as it gets!" - a proposal for a new flight simulator - REVISION 2.0 ---- by David Murr (dmurr@pen.k12.va.us) -------------------------------------------------------------------- ** Please note that this file is a revision from an earlier on that was sent out. It reflects some of the thoughts and observations made by members of the development team, and of the "public". ** I. An introduction to the project An idea entered my head on Monday, April 8, 1996. "'Why wait around for some large company to make what WE want? Why not do it ourselves'". Think about it -- there's certainly not a lack creativty, ingenuity, or desire among our tight-knit group of about 1 000. What have we been posting to newsgroups for months, even years? Wish lists. We forward them to webmasters at Microsoft, Sierra, SubLogic, whomever, only to have the e-mails dumped in the "recycling bin". Let's examine what we have out there -- what has been done already with what huge corporations like Microsoft have given us. 1) We have dissected the file structure of nearly EVERYTHING to some extent with MS Flight Simulator 5.x. 2) Maurizio Gavoli (please excuse my spelling) and someone else whose name escapes me at the moment have created two excellent aircraft for MSFS5 - without the need of any add-on programs! 3) Chuck Dome, who "reverse-engineered" the .AIRcraft files and came up with AIRMOD, an excellent utility used to modify flight and color characteristics of default (and the two Simula FlightWare planes) aircraft for FS5. Thanks, Chuck for a utility that certainly tided us over until the release of Flight Shop. 4) Curt Olson's Real-Wx, which allows the user to download real, current weather for a multitude of US and Canadian cities and automatically import that data directly into FS5! 5) The creators of FSASM, SCASM, BGLCOMP, BGLGEN, Airport, and countless other scenery creation utilities that have undone the .bgl files and made the format publicly available, besides giving us excellent tools to make our own, even photo-realitic, scenery! ... and countless other, excellent contributions by generous simmers! (Sorry guys, I don't have the space or time to list you all, but you know who you are!) What does this have to do with Flight Gear, you ask? Well, I'm sitting here, wondering, "With all this creativeness, WHY wouldn't be possible to make our own simulator?". It only makes sense - it's the next logical progression. Isn't it? So, now that you know where I'm coming from, let's get to some of the specifics of the program. II. Overview of main features and specs Here is a "condensed" (though lengthy) list of things we DEFINATELY need in this program: * Interactive, "smart" ATC that SPEAKS when there is a sound card present Perhaps two-way, verbal communcation between ATC and pilot. It would also know the topography of the scenery area so it doesn't "lead you into a mountain"! * A "CUSeeMe" type communication to other pilots has been proposed, however this would take a tremendous amount of bandwidth and is not feasible at THIS POINT. A more reasonable approach would be to perhaps devote a server (we don't want to overload Curt's wx server now do we? ;-) to providing a few "radio frequencies" (not really radio, of course) and using those, so you can talk to an automated "intelligent" ATC and other pilots on that frequency! Sounds fun, huh? :) * ATP-level or better graphics (hopefully better than FS5, though) * IT MIGHT BE POSSIBLE TO RUN FG ON A 386! (Notice I said *POSSIBLE*, but probably not FEASIBLE!) * Acceptable frame rate (~10fps) on a 486DX/50 ** On a future possible standard 80686/250MHz computer, let's aim for no less than a cinematic frame rate, i.e., 25fps minimum. * Probable minimum setup: 486SX/25 with 8-12 megs of RAM, 1MB VRAM, 40MB of harddrive space plus 10 MB or more for aircraft, scenery, etc. * 32-bit under DOS with a DMPI extender * Written in C (GNU/DJGPP compilers), Ada, and assembly. * Different, portable (both meanings) modules for each sections (e.g.: scenery, wx, flight dyn, etc.) ** Optimum system at current time: - 80586 (Pentium) DX/120MHz processor - Math Coprocessor - DOS 6.2 or later - 12-16 megs of RAM - 1-2 MB of VRAM (video card's memory) - graphics accelarator board - Full-time, 56.6Kbps ISDN Network Connection (Internet) However, a minimum dial-in 14.4Kbps account is acceptable. ** NOTE: this is only used for those wishing to get wx, scenery, and ATC updates on the fly from a centralized Internet server. (Location to be determined) - A DEDICATED FLIGHT PROCESSOR! (Though not at all neccesary, it will make your flights a lot smoother if you have a dedicated computer connected to the main one that handles the network updates. (One of the main advantages hinges on connectivity, though the user will be able to enjoy many enhancements over current simulators with advanced polygon-rendering routines, 32-bit architecture, open documentation of the file structures to encourage 3rd party FREEWARE addons... note: no shareware will be allowed to be associated with this program!) - A Sound Blaster compatible 16bit sound card (full duplex) for voice communcation with other pilots/ATC. - SVGA 800x600, 24bit color - What else? * NO LONGER COMPATIBLE W/FS5 FILES! (??) Undecided at this point * Well documented proprietory format for a/c, scenery, and panels * Accurate, flyable flight models * Accurate, but non-processor intensive graphics (good frame rate) No "bitmap textures". Remember the chunky textures in FS5 when you're lower than ~4000ft? None of those, hopefully. * Accurate REAL WORLD representation of objects, especially major landmarks and monuments. * Simple on-board radar system with GPS "moving map" display * Default planes include Cessna SkyLane RG, Beechcraft Bonanza, a hot air balloon, Boeing 777, maybe a float plane? What else?? * Ability to construct abstract shapes easily for scenery, such as the St. Louis arch, Mt. Rushmore, Leaning Tower of Pisa, etc. * Written in C(++), assembly, and Ada? for controlled, robust, fast code for optimum performance. * BONUS: We'll probably port it to Unix, Mac, and OS/2! * RUNS UNDER DOS 5.0 OR ABOVE! NO CD's AND **NO** WINDOWS 95! YAY! * FREE! Flight Gear MUST be FREEWARE and easily accessible via the Internet, AOL, Prodigy, CompuServe, Wow!, etc. * The source code will be available for free, plus extensive docs on how we constructed different modules, etc. * Some tech support by e-mail in spare time. No guarantees on this. ** Since we're talking about a possibly 5-year project (though estimates have ranged from 2-10 years! :), our speculations on system requirements will most likely be WAY OFF, but since we don't know what's going to happen to computers and networking in the future, and thus can't plan that far ahead, we'll just have to adjust our code accordingly when it is deemed necessary. ** **** GNU C/C++ standards will be used in C development. **** Curt Olson proposes we use a centralized server for wx, scenery, etc. that IS REAL TIME for those that are connected to the Internet at the time of program initialization. This would work as such: when the user starts Flight Gear, it will test to see if a connection is available, and if so, it will contact the wx server and download current weather information and implement that in the sim, as well as polling the server at a specific interval, say, 10 minutes to download currently conditions... Perhaps we could also tie in with a real-life ATC and re-create REAL AIR TRAFFIC!! Automatic scenery updates would be available via this server as well. For those who are not connected, we could implement a REALWX type of program or just use RANDWXGEN...? Maurizio suggests we come up with a better mechanism for creating scenery than BAO's. He says that the way the .bgl files are put together is very wasteful and slow. How can we render the buildings, scenery faster...? How do we get realistic-looking scenery without those framerate-killing, awful bitmaps? They're chunky below 4000ft, and they are too SLOW! I will continue to put more into here as warranted. Thanks. /dlm III. So, who's gonna put this thing together? Good question. Here is a list of people I *HOPE* to get on our team. Keep in mind that I haven't contacted some of them yet, so... * Maurizio Gavioli and Chuck Dome - implementation of aircraft * Curt Olson - STN and wx development, C aide * Enno Borgsteede and the SCASM/FSASM development teams - scenery * Mark Stotzer - project consultant/supervisor * Mike Vidal - flight modeling (MG has signed on, as well as Michal Rok and CO. We have others as well, but we're just getting ourselves together at this point. We hope to get some "outside" help from C/ASM experts.) ... and others. These people seem to be experts in the particular field I have placed them in, so... IV. Got any kind of release date yet? Let's say "Years." As in five years. Who knows? V. Miscellaneous Got any input for this project? THEN PLEASE e-mail me at dmurr@pen.k12.va.us! Same goes if you want to volunteer your time/resources. Anything you'd like to add? Please let me know -- my e-mail address is: dmurr@pen.k12.va.us. Oh, BTW, supplemental info will be available at: http://users.aol.com/fsnews/flightgear/. If you want to talk to all the developers at one time, send a message to: fl182@cleveland.freenet.edu -- that's our mailing list to discuss things openly. If you're not on the design team, please keep your messages short as we are constantly shipping lengthy msg's back and forth among the team. Regards, David Murr Editor - FSNews http://users.aol.com/fsnews/flightgear/ <--- Coming soon! Updates on Web! I WILL POST REVISIONS OF THIS DOCUMENT ON R.A.S. AND THE IUP MAILING LIST AS WARRANTED.