I've heard Flight Gear enthusiasts claim that the FG scenery already exceeds the quality of other flight sims in terms of accuracy, and I would believe it considering how large the scenery files are (isn't it like 3 DVD's for the full version with all scenery?). What I don't understand is what is included in all this data that the scenery looks so FS98-ish?For instance, for scenery files that large, I would expect rivers to have more well defined coast lines, cities and forests to have better defined and more varietied land classes, and certainly more realistic and more realisticly meshing textures.There are a few freeware texture developers, and certainly some talented other folks out there... has anyone thought of starting a fund for paying someone very talented with graphics to develop a set of nice-looking textures for Flight Gear? I would certainly donate to such a fund. Not to knock the cause of the website (or make myself seem insensitive), but I think I'd rather donate my money to FlightGear and know it went to that than go to sponsoring a child. I can sponsor a child separately, on my own.Similarly, has anyone thought of starting a bounty fund for the development of a GPS system, or of more advanced ATC? I know that FlightGear doesn't necessarily have well defined goals, but if things were a little more organized and focused it would seem to have more appeal.I know, I fully realize that I just inserted my foot into my mouth... I don't have the technical ability or time to donate to FlightGear, and the developers are doing what they are doing out of love alone. I deeply respect that and thank the contributers for the progress that has been made. I was just curious.Any news of a 0.9.9 or what features may or may not be included for the big 1.0 release? Have the website maintainers ever thought of adding RSS feeds to the website?Ken

>I've heard Flight Gear enthusiasts claim that the FG scenery>already exceeds the quality of other flight sims in terms of>accuracy, and I would believe it considering how large the>scenery files are This is true for the mesh when comparing with FS2004 and was truefor the mesh when comparing with X-Plane (7.x and lower).FS2004 uses only low resolution mesh data for the standard scenery,Flightgear uses high resolution data from the SRTM space shuttle mission with a resolution up to 3 arcsec, that's about 90 m.Sure you can buy addons for FS2004 that are based on the SRTM Data too, but FS2004 doesn't use this data very efficientlybecause the 3d engine of FS2004 can only handle a regular mesh.A Regular mesh tends to be unrealistic, because you can only display triangles with angles of 45 and 90 degrees (when seen from top).Also the distance between one vertex to the next is allways regular which does not look natural.This means a distance from 1, 1/2 or 1/4 is possible, but not 1/3, 0,4762 or 0.13 etc .And the worst of all this is, a regular mesh is not efficent,it needs to much computer power compared to an irregular mesh.Because you need to display and calculate every vertex point of your regular mesh, even when this vertex pointhas the same high as its two neighbours. That's for example the case on the sea or a salt lake.All this is not the case for an irregular mesh which is used by Flightgear and now X-Plane 8.x.Here a triangle can have every kind of angle, so a scenery does look much more natural.An irregular mesh also saves computer power, because an irregalur mesh doesn't need to calculate unnecessary vertex points.Here are two example screenshots of an irregular mesh:http://www.flightgear.org/Gallery.hist01/L...irregular1.htmlhttp://www.flightgear.org/Gallery.hist01/L...irregular2.htmlFor comparission, a regular mesh looks like a chess board.> (isn't it like 3 DVD's for the full version>with all scenery?). Yes, 3 DVDs, that's about 12 GB data:http://cdrom.flightgear.org/> What I don't understand is what is>included in all this data that the scenery looks so FS98-ish?The scenery in FS98 is flat with hills that look like pyramids set on top of the flat ground.The scenery of FlightGear is never flat, even the sea is not flat because in FlightGear the earth is round and this can be seen in flightgear from high altitudes.In FS2004 the earth is not round, it is only a cylinder withbarriers at the north and south pole which can't be crossed.So what you mean with "looks so FS98-ish" are definatly only the textures.Here the problem is, that Flightgear does only use single textures for the ground at the moment because there is no support for multitexturing in the plib library which flightgear is using as scence graph library.To change that, the plib library must be improved first,then you can use this feature in FlightGear too.The plib library is the big problem:plib.sf.net>>For instance, for scenery files that large, I would expect>rivers to have more well defined coast lines, cities and>forests to have better defined and more varietied land>classes, and certainly more realistic and more realisticly>meshing textures.Well the scenery depends on the real world data you can get.And flightgear can only use free real world data becausedata that costs money is expensive and in most times incompatible with the GPL license which is used by flightgear.The problem is, there is not much free real world data available.Flightgear does allready use in most cases the best data you can get for free.The other problem is, the real world data is not perfectand there are many reasons why this is the case, for example low resolution measuring and things like that.But the varieties of land classed could be improved, but this is a big task. Volunteers are needed.>There are a few freeware texture developers, and certainly>some talented other folks out there... has anyone thought of>starting a fund for paying someone very talented with graphics>to develop a set of nice-looking textures for Flight Gear?Textures are not the problem.The landclass data that defines where at which location what kind of texture or land cover must be set is the problem.But from a graphic artist standpoint FlightGear could need more modelers who create 3d models of buildings and things like that for Flightgear.This is at the moment the easiest way to improve the overall scenery in FlightGear. It would also be nice to have some sound artists or volunteers who record sound effects from real airplanes for FlightGear.By doing this, keep in mind, all data must be compliant with the rules in the GPL license.http://www.gnu.org/licenses/gpl.htmlhttp://en.wikipedia.org/wiki/GPL>>Similarly, has anyone thought of starting a bounty fund for>the development of a GPS system, or of more advanced ATC? I>know that FlightGear doesn't necessarily have well defined>goals, but if things were a little more organized and focused>it would seem to have more appeal.There are plans to develop this, but time is an issue.Paying a flightgear developer for working full time on FlightGear would help a lot. :)>Any news of a 0.9.9 or what features may or may not be>included for the big 1.0 release? All new features are available in the cvs version of flightgearbut i don't know what new features get activated in the 0.9.9 stable release. I estimate features like the new 3d clouds or the AI and some new planes, but i am not sure.

Hi,I think you are bringing up a number of valid points, however they're not going to be very popular,also the FlightGear mailing list (devel) would probably be a better place to address theseissues, anyway:>I've heard Flight Gear enthusiasts claim that the FG scenery>already exceeds the quality of other flight sims in terms of>accuracy, and I would believe it considering how large the>scenery files are (isn't it like 3 DVD's for the full version>with all scenery?). What I don't understand is what is>included in all this data that the scenery looks so FS98-ish?Apart from what's been pointed out by the previous poster, the majorproblem seems to be the approach how scenery and terrain generation are handled by Flightgear.That is, while the underlying terrain data can provide much moredetails than some of FlightGear's commercial "competitors", implementations of other FlightGear components are currently way too static, and thereby restrict the true potential of the simulatorinherently, i.e. the scenery is currently also practically exclusively statically generated.>For instance, for scenery files that large, I would expect>rivers to have more well defined coast lines, cities and>forests to have better defined and more varietied land>classes, and certainly more realistic and more realisticly>meshing textures.The scenery files themselves don't even need to be that large if thescenery itself were created dynamically, however that's quite a task.Even though some other (commerical) flight simulators do already featuresimilar functionality.But currently the scenery is statically created, which also explains thelack of scenery data for the majority of areas in FlightGear.So, with the exception of those areas that were populated with objectsby volunteer scenery designers (i.e. San Fransisco area), the majority of areas are going to look rather plain and simple - and mostly even inferior to FS98 scenery.Hence, in order to truly enable FlightGear's scenery/terrain engine tobecome as flexible as most of FlightGear's other components, this partof FlightGear would need to be re-designed carefully, so that terrainand scenery are indeed appropriately separated, but can be linked easily.For example much like X-Plane handles scenery, where airports (runways/navaids)will also be properly generated even without having installed any addon scenery.So, basically this would also mean to enable FlightGear to create scenery objects "on the fly", based on pre-determined rules, i.e. so that the actual data is retrieved and parsed from ascii/xml files.Likewise, coast lines, continents etc. would no longer be statically defined byusing pre generated scenery files, but could rather dynamically be adjusted, thatway FlightGear would would be able to natively deal with most of the data, ratherthan using separately generated scenery for the basic (and most common) scenery objects.After all, this would also make FlightGear much more appealing to the average userwho would otherwise still have to download huge chunks of so called "scenery files",that often do not contain much "scenery" at all.

>There are a few freeware texture developers, and certainly>some talented other folks out there... has anyone thought of>starting a fund for paying someone very talented with graphics>to develop a set of nice-looking textures for Flight Gear?I don't think that's going to happen either, apart from the factthat said money would first need to be made, none of the coredevelopers is probably going to actually think about basicallygiving money to others to make them contribute to the project. Which is indeed pretty understandable, as these guys usually have spent years contributing to the project without any reimbursement at all, usually even having to spend huge amounts of money for their contributions (i.e. servers, bandwidth, hardware etc).Essentially, you are suggesting to "outsource" parts of thedevelopment of FlightGear, which is certainly not going toappeal to the majority of developers anyway, and which wouldotherwise also require corresponding funding, funding which simply isn't available.Also, I'm pretty sure that if there was some more "PR" going on, the project could probably easily attract more volunteers for either tasks, be it texture design, scenery design, aircraft design or actually C++ coding.So, personally I don't see the need to actually payanybody to come up with better textures either, there are certainlyother components of FlightGear that need more attention.

>I would certainly donate to such a fund.You might find it enlightening to look up on the mailing list archives,there were some lengthy discussions about donations and the like a whileago if I remember correctly.Personally, I'd also consider donating money to the project (could be PayPal based) in order to push development of certain features/components that I consider important, and I know other FlightGear users who'd do that as well if there was the possibility.I even think that this specific possibility has also been brought up some time ago on the devel list, but basically the idea was also rejected, if I remember correctly because of some people being worried about potential tax issues that might be created if the FlightGear project accepted donations.Even though someone considered to set up a non profit organization forFlightGear.I'm not sure how involved this would really be, but I doubt that it canultimately be such an issue because various other opensource projectsare making use of community based donations, meanwhile even SourceForgefeatures the possibility to make donations to a project, so theoriticallyit should be pretty trivial to set up such a possibility.However, you are checking the mailing list archives, you'll also notice thatthey even had a hard time to even consider putting google ads on flightgear.orgSo, they are certainly not very market-oriented.But I agree that for people like me, who are not so much in the positionto actually contribute much by coding because of their full time job, donationswould actually be a good way to contribute to the project nevertheless.

>Not to knock the cause>of the website (or make myself seem insensitive), but I think>I'd rather donate my money to FlightGear and know it went to>that than go to sponsoring a child. I can sponsor a child>separately, on my own.Well, kinda courageous to actually mention this publicly.Admittedly, I have also talked to two other FlightGear users who think similarly and do also consider this a potentially controversial way to deal with potential revenue that is actually preferably intended to be used to make the project progress.Currently, as someone who considers to support FlightGear financially (i.e. by purchasing a set of DVDs) you basically need to realize that any financial contribution is unlikely to be actually used for the sake of the project, which is somewhat sad - for the project.On the other hand this is really difficult to discuss, after all this is an absolutely noble purpose and it's difficult to suggest directing potential financial contributions/revenue towards FlightGear DEVELOPMENT rather than using it for humanitarian purposes.I'm pretty sure that there are several people out there who are having a similarly difficult time thinking about how to actually bring thisissue up without sounding absolutely insensitive, given the relative wealth and luxury that most of us experience each day compared to thepeople living in the 3rd world.Indeed, I did initially also hesitate to actually comment on your remark concerning this issue - for good reasons.Around christmas however, I found it quite noble to read that the FlightGear project donated some money to the Tsunami victims, but it's clear that this money cannot be used for FlightGear development anymore.Taking this into account, as well as the repeated public reminders of FlightGear developers and contributors on the FlightGear Devel mailing list that they are not paid for their work, but would definitely like to be paid, it's hard to understand how this is supposed toactually work out without some real changes getting considered.So, this in itself is also some sort of vicious circle: If the project was more known and popular, it could probably easily attract more potential developers and users - which would probably also result in some more funds getting in, whereas some of these could still be used for whatever noble purposes the project coordinator deems suitable, the majority could benefit the project itself to ensure that it actually grows at a steady pace.However, in order to actually get more known and popular, the FlightGear project would have to actually get some things sorted out in the first place, so that FlightGear indeed becomes easily usable for the average user - which is unlikely to happen anytime soon anyway.Unfortunately, for the average user, even MS FS98 remains usually a much more feasible option than using FlightGear,no matter whether that user wants to use FlightGear as a game or as a training aid. This is another thing that needsto be taken into account: the majority of community contributions for MS FS (scenery, aircraft, adventures etc.) comefrom the GAMING community, not generally from folks who want to use the simulator scientifically or even professionally.So, FlightGear requires currently a rather steep learning curve or corresponding background in order to actually put it to good use, something which one cannot expect from the ordinary user.However, ultimately it would of course be inacceptable to suddenly stop providing any sort of continuous humanitariansupport if indeed a child relies on it meanwhile, so this would be another thing to take into account when/if things arereconsidered.

>Even though someone considered to set up a non profit>organization for FlightGear.Actually, a non profit organization would definitely have many advantages, i.e. the majority of big opensource projects seems to take that route in order to easily deal with donations etc.Also, for some opensource competitions this seems to be some sort of inofficial requirement, i.e. the majority of all mentoring organizations that are participating in Google's "Summer of Code" program seem to be NON PROFIT organizations, too (http://code.google.com/summerofcode.html).The whole SOC thing is a very interesting thing anyway, think about the money each developer ($4500 US) and the project itself ($500) could make.

I like your posts, hfitz. They are very detailed and informative.In addition to what you have suggested, I think there are two other ways for FlightGear to get more fundings. The first would be to sell some CD's on aircrafts. Unfortunately, none of the aicrafts are truly in production status right now. There are a few that are, but they are from WWI-era, and are not too interesting to the average simmer. The second would be to sell some CD's on local sceneries. Unfortunately, beside the San Francsico area, there is no other scenery to sell.If more efforts can be put into these two areas, not only will there be more fundings, but there will also be more people attracted to FlightGear.

>>Even though someone considered to set up a non profit>>organization for FlightGear.>>Actually, a non profit organization would definitely have many>advantages, i.e. the majority of big opensource projects seems>to take that route in order to easily deal with donations>etc.>>Also, for some opensource competitions this seems to be some>sort of inofficial requirement, i.e. the majority of all>mentoring organizations that are participating in Google's>"Summer of Code" program seem to be NON PROFIT organizations,>too (http://code.google.com/summerofcode.html).>>The whole SOC thing is a very interesting thing anyway, think>about the money each developer ($4500 US) and the project>itself ($500) could make.Google just started accepting new organizations/projects for its upcoming summer of code:http://code.google.com/summerofcode.htmlMaybe this would be interesting for FlightGear, too?Of course, this would probably require a collection of various FlightGear related projects, or at least some sort of bug/feature tracker.