Advanced Weather v1.4 in Flightgear 2.6+

Part I: Convective clouds

Author: Thorsten Renk

The screenshots shown in the following use shaders, textures and scenery which are for various reasons (incompatible license, too recent development,…) not part of the official Flightgear 2.6 release. However, these are available for download and every feature works with Flightgear 2.6. The following packages need to be installed in addition to get to see the same: lightfield shader package v1.1, Juneau custom scenery, and textures from regional textures v0.1.

An integrated weather system

Without a doubt, clouds, haze and fog are the most easily noticed features of weather in a flight simulation, followed by winds. Advanced Weather v1.4 is however more than a tool to draw clouds and set wind parameters – it is a system with a (limited) understanding what the weather which it currently tries to render is, and it aims to simulate features of atmospheric physics.

This means that different weather phenomena tie together – winds and terrain influence the way cloud formation is taking place, cloud formation and the formation of thermal updrafts is connected, and weather is always understood as part of a large-scale weather pattern involving high and low pressure systems.

At the same time, clouds and atmospheric haze also influence the atmospheric light (and now also the scenery) in an essential way – strong fogging changes the color of sunrises to a blue-grey, wave patterns on the ocean follow the wind strength and direction and rain causes visibly wet runways. Let us have a look at how this works in fair weather.

The formation of convective clouds

Fair weather is typically characterized by convective cloud development: The sun heats the terrain and the air layer just above, thus warm air rises up in ‘bubbles’ and forms thermals, as the air rises, it expands and cools and eventually the moisture condenses into droplets, forming the characteristic, cauliflower-shaped Cumulus clouds. Cumulus clouds are the most common example of clouds formed by pronounced vertical motion of air.

As every glider pilot looking for thermals has to learn quickly, the formation of convective clouds depends on many different factors. The terrain type is crucial – while rock or concrete surfaces heat well in sunshine and may easily lead to well-developed thermals and cloud formation, open water or ice is much less likely to heat up in sunshine and seed Cumulus formation. High points in the terrain mark the spots where the bubbles of warm air are most likely to lift off the ground. Another important factor is the time of the day: The sun needs sufficient time to heat the terrain, therefore Cumulus formation is densest around noon, but the thermal updrafts are strongest in the afternoon, and while pronounced Cumulus clouds are unlikely to form in the morning, the thermal energy accumulated over a day may still give rise to well-developed clouds in the evening.

Let’s follow the development of convective clouds during a day in Juneau (Alaska). At sunrise, only very few clouds form, and they are transient, whispy phenomena (click to enlarge images):

Later in the day, the cloud formation is somewhat stronger. Note how clouds tend to form over mountain peaks, but do not form over open water. Also, no strong cloud development occurs over Taku glacier in the upper left, despite its high altitude, as the ice surface does not heat well in the sun.

At noon, the thermal updrafts become stronger and consequently the clouds become more well-defined. While in the morning the upward motion of air rarely exceeds 0.5 m/s, around noon this becomes rather 1 m/s, to strengthen even more in the afternoon.

Yet a few hours later, the number of clouds decreases again as the thermal irradiation by the sun weakens, but then typically fewer but stronger thermals with larger cap clouds are found.

Towards sunset, there is still significant thermal energy left to lead to sizeable cloud development, although the number of clouds as well as the typical strength of thermals is decreasing again. During the night, the development of convective clouds breaks more or less down completely as there is no thermal energy from the sun available.

Interaction of convective clouds, wind and the terrain

Wind meeting a terrain barrier corresponds to an upward-moving airflow, and hence is able to alter the development pattern of convective clouds in an essential way. Consider the following scene above Maui (Hawaii) in the absence of winds. Clouds rim the peak of Haleakala, but do not actually reach all the way up to the mountaintop (note also that due to the different latitude of Hawaii, there is far more thermal energy coming from the sun than in Juneau, leading to a much higher overall density of clouds):

With 20 kt winds coming from the north, the picture changes quite drastically: clouds are now pushed up all the way to the summit of Haleakala by the rising air, whereas the falling air south of the crater creates a lee effect in which convective clouds disappear. The vegetation pattern of Maui reflects these prevaling conditions – while moist air is carried up all the way to the summit of Haleakala, it rains off and irrigates the northern slopes of the mountain, leading to a bright green forest belt. The southern slopes on the other hand see typically falling air and dissolving clouds, and are consequently much drier.

Convective clouds in flight

If the appropriate option is selected, thermals are automatically generated along with Cumulus clouds so that thermal soaring is possible. Combined with the effect of ridge lift, this can make for rather realistic mountain soaring conditions in which a good degree of skill is required to stay in the air.

But the detailed interplay between convection and the terrain leads to interesting scenes also in other planes. Around noon, the peaks of high mountains are often covered in rather dramatic clouds piling up against the slopes.

Further down, the altitude of the cloudbase is no longer determined by the terrain but by the air layers.

And yet, the terrain elevation and the change between land and water imprint a pronounced pattern onto the distribution of density, shape and size of convective clouds.

Convection may also occur due to vertical instabilities in upper air layers, leading to the development of Altocumulus clouds, or at even higher altitudes Cirrocumulus clouds. Here’s an example of the development of Altocumulus fields at 15.000 ft. At such high altitudes, the clouds are no longer influenced by the terrain underneath, but rather by the properties of the air layers between which the Altocumuli form. For instance, Altocumulus development may be caused by the instabilities associated with an approaching cold front, and may thus signal the danger of severe thunderstorms in the near future. The Advanced Weather offline weather generation automatically includes this and other rules of large-scale weather patterns.

Next: Layered clouds, haze, fog and precipitation

FlightGear v2.6.0 Released

Also available in: NederlandsEspañolFrançaisPortuguêsRomână


The FlightGear development team is happy to announce the v2.6.0 release of FlightGear, the free, open-source flight simulator. This new version contains many exciting new features, enhancements and bugfixes. Major improvements from v2.4.0 include reduced AI aircraft load times, easier graphics tuning, more sophisticated AI aircraft and improved usability.

Founded in 1997, FlightGear is developed by a worldwide group of volunteers, brought together by a shared ambition to create the most realistic flight simulator possible that is free to use, modify and distribute. FlightGear is used all over the world by desktop flight simulator enthusiasts, for research in Universities and for interactive exhibits in museums.

FlightGear features more than 400 aircraft, a worldwide scenery database, a worldwide multi-player environment, detailed sky modelling, a flexible and open aircraft modelling system, varied networking options, multiple display support, a powerful scripting language and an open architecture. Best of all, being open-source, the simulator is owned by the community and everyone is encouraged to contribute.

Start downloading FlightGear v2.6.0 for free from

FlightGear – Fly Free!

Some of the major changes include:

Aircraft operations:

  • At selected airports, FlightGear can automatically start at an appropriate parking spot based on the size and type of your aircraft.
  • At airports which support this feature, a visual display of the taxi route on the ground guides you to the active runway, while following the correct taxiways.

AI system

  • To reduce stuttering during model loading and take further advantage of multi-core CPUs, MP and AI aircraft models are now loaded in a background thread.
  • To reduce load times still further, only the parts of the aircraft currently visible are loaded from disk.
  • AI and multiplayer aircraft are no longer silent objects, they can produce sounds just like the main aircraft.

AI Traffic

  • Many new and updated AI aircraft and liveries. Over 80 airlines now populate the virtual skies.
  • The range at which AI aircraft are displayed is now configurable, allowing you to tune FlightGear for best performance.
  • AI controlled pilots have received extensive landing training and now make a more realistic approach and vacate the runway when able.
  • The simulator now assigns an available parking position on startup.

Flight dynamics

  • The JSBSim flight dynamics model received a major overhaul.


  • The Local Weather package has been further integrated with the FlightGear core, and has been renamed “Advanced Weather”. New rendering techniques allow more detailed clouds with no performance impact. High altitude clouds are rendered more realistically, and clouds move with the wind without impacting performance.


  • New replay system. Video-player like controls, including slow motion and fast forward. Pilots can now take over the aircraft at any time during a replay. Great for training particular flight phases such as approach over and over again.
  • Aircraft status ratings are displayed in the FlightGear launcher, allowing you to see at a glance the FDM, model, systems and cockpit quality for rated aircraft.
  • Multiplayer settings can be accessed in-sim. You can now choose your callsign, select an MP server and connect within the simulator.
  • Automatic scenery download is now even easier to use. Simply select the scenery directory to download to, and switch it on.
  • Individual graphics effects can now be configured from within the Rendering Settings dialog, allowing you to fine-tune the performance of FlightGear within the sim.
  • The simulation of radio signal propagation has started and will make the reception of ATC messages and navigation aids more realistic in the future.
  • A new set of options makes it easier to create seamless and zoomable multi screen setups.
  • A new performance monitor shows the time spent in each subsystem.

Highlighted new and improved aircraft

Project infrastructure

  • CMake is now the official build system on Linux, Mac and Windows.

Visual effects

  • The sea now looks more realistic. Waves align with the wind, and foam appears at high wind speed.
  • Steep slopes now appear rocky.
  • Runways now appear wet during rain showers.
  • To help aircraft developers, a single shader combining bump-map, specular, reflection and light mapping components is now available.


  • Additional joysticks and rudder pedals are supported out-of-the-box, including the Logitech WingMan Interceptor, Saitek Pro Combat Rudder Pedals and Thrustmaster HOTAS Warthog.
  • FGPanel, lightweight software to render 2D instrument panels, is now included as part of the release.

Bug fixes

  • See our bugtracker for an extensive list of the bugs fixed in this release.


  • 64bit version restored and its installation integrated in the main fgsetup installer

Sky Diving Visualization

The Challenge

As a skydiver adds more gear such as front packs and items strapped to legs or arms, the jumper’s basic stability in free-fall is reduced.  It becomes easier to tumble out of control and there is less margin for error.  Similarly, the aerodynamic wake of the jumper may interfere with pilot chute opening (known as “hesitation”). Investigating different gear configurations generally involves vertical wind tunnel testing, or actual tests with jumpers. To avoid some of the cost, and mitigating safety concerns, a tool to computationally analyze these jumpers and their gear is highly desired. Creare, Inc., an R+D research firm in Hanover, NH, under funding from the US Army, developed a Computation Fluid Dynamics toolkit for analyzing jumpers and their equipment, and model the resulting configurations in FlightGear.

Note: a flyable parachutist model is available to download and test at the end of this article.

Fluent (CFD) and Stability Derivatives

CFD = Computational Fluid Dynamics.  ANSYS Fluent is a high end CFD that models flow, turbulence, and heat transfer in 3d.   Imagine being able to take a 3d model of a sky diver (or an aircraft) and place it at different orientations and different poses.  Then for each orientation and pose, run a computer simulation of exactly how the air flows around the sky diver, where pockets of turbulence are generated, and what forces and moments are produced.  Imagine all the combinations of roll and pitch and body poses possible — it leads to a huge number of combinations.  Now imagine repeating that for several different arrangements of front and back packs and other equipment.  You will need a cluster of computers running for days or even weeks to compute all the permutations just for a single pack configuration.  This is essentially a “virtual wind tunnel” running on a super computer cluster of PC’s.

In the case of the parachutist simulation: the amount of computation time required to generate 2 scenarios (with a back pack and without) was approximately 25,000 cpu-hours — or around 2 years of compute time on a single processor PC.  1000 individual simulations were run, each involving approximately 4 million “elements”.

One of the detailed CFD models used by Creare



Real world testing and data collection was performed in a vertical wind tunnel (such as the one linked here.)  This real world data could then be compared to the the Fluent (CFD) results to validate and possibly improve the computer model.

Real Time Simulation

Build-Up of Coefficients:

  • For each component of the model, the local angle of attack and sideslip angle are calculated from the combination of the limb orientation and the overall angle of attack and sideslip of the entire model.
  • For each of the six degrees of freedom, the contribution of the model component to the overall aerodynamic response is calculated from tables of non-dimensional coefficients:
    CFx,y,z = Fx,y,z / ( q * Acomp )
    CMx,y,z = Mx,y,z / ( q * Acomp * Lcomp )
  • The two-dimensional lookup tables are compiled from data extracted from the CFD results in the Solution Database.

Forces and Moments:

  • Aerodynamic forces and moments are transformed from the local frame to the global frame and then summed.
  • Resultant forces and moments then determine accelerations, velocity and turn rates are calculated, and the model iterates.

Creare partnered with Jon Berndt (the founder of JSBSim–one of the core physics engines used by FlightGear) to contribute some clever additions to JSBSim that permit a “blade element” approach to parachutist modeling.  Jon helped tremendously optimizing and integrating the required new code into JSBSim which then ultimately led to its inclusion in FlightGear.  The parachutist physics model is an order of magnitude more complex than a typical aircraft model.

Figure Animation and Posing

The character model is built out of several animated subcomponents: left & right forearms, left & right upper arms, left & right lower legs, left & right upper legs, head, torso, and pelvis.  The model parts are attached in a cascading fashion like a real figure, and each joint can be rotated through all 3 axis (roll, pitch, and yaw.)  In order to avoid unrealistic contortions, sensible joint range of motions are defined.

There are a number of predefined poses where the appropriate joint angles have all been worked out in advance.

  • Box: a neutral pose minimizing rotational or translational motion.
  • Left & Right Translation: mirrored poses that induce a “slide” either to the left or right.
  • Anterior Translation: a pose that induces a forward slide.
  • Posterior Translation: a pose that induces a rearward slide.
  • Left & Right Dorsoventral: mirrored poses that induce a left or right rotation (yaw.)
  • Dorsal: a “spread eagle” pose that maximizes surface area and thus minimizes decent rate.
  • Ventral: a “compressed” pose that minimizes surface area thus maximizes decent rate.
You might notice that these poses map rather neatly into well understood pilot controls similar to flying a helicopter.  For the FlightGear simulation we can mix these poses together in proportion to the corresponding joystick axis deflection and throttle position and fly the sky-diver intuitively.  For those that doubt, this actually works quite well! 🙂

Visualizing CFD Flow-lines

One of the neat things that a CFD analysis can produce are airflow lines that pass around the model.  We can take the 3d flowlines that are produced by the CFD and attach them to the 3d model of the figure.  This allows visualizing the flow lines from any FlightGear perspective.  One interesting technical challenge is that the flow lines need to keep a fixed vertical orientation even though the model may roll or pitch, yet the flowlines must track the heading/yaw of the model.  This can be done by setting up appropriate inverse transformations in the FlightGear model animation configuration file.

Smoke and Trajectory Markers

FlightGear offers additional visualization aids.  The model is set up to support emitting smoke.  FlightGear smoke drifts with the prevaling winds (which can often be substantial at higher altitudes.)  The model is also setup to emit “trajectory markers” at a fixed rate.  The trajectory markers stay fixed in 3d space and represent the actual path the sky diver follows.  In addition they represent the orientation of the sky diver at that point in space.

Where is the Parachute?

This exercise is setup as a free-fall simulation, not a parachute simulation so there is no chute modeled.  Instead the simulation is mercifully paused when the altitude reaches 100′ above the surface.

Download and Fly

Follow these instructions to download, install, and fly the Creare Parachutist model:

  • Note: the parachutist model is not compatible with FlightGear v2.4, you must fly this model with one of the v2.6 release candidates, or the official v2.6 release scheduled for February 17.
  • Download the file.
  • Unzip it into your FlightGear “Aircraft” folder.
  • Start FlightGear and select either –aircraft=Parachutist-Scenario1 or –aircraft=Parachutist-Scenario2
  • Make sure you specify  an initial altitude (such as –altitude=10000), otherwise you will just be sitting at the end of the runway working on your tan.
  • Press F1 and F2 to toggle the two available dialog boxes on/off.
  • You can manipulate the joint poses individually or select from a set of pre-defined poses, or select “Joystick” control and fly with a joystick (or keyboard or mouse) similar to flying a helicopter or airplane.


  • Dietz, A. J., Kaszeta, R. W., Cameron, B., Micka, D. J., Deserranno, D. and Craley, J.”A CFD Toolkit for Modeling Parachutists in Freefall”, presented at the 21st AIAA Decelerators Conference in Dublin, Ireland, 23-26 May 2011. Paper AIAA 2011-2589.
  • The Creare Freefalling Parachutist model was developed by Anthony Dietz (Principal Investigator), Richard Kaszeta , Benjamin Cameron, Daniel Micka, and Dimitri Deserranno of Creare, Inc., as part of an SBIR Phase II project sponsored by the U.S. Army RDECOM Acquisition Center under Contract No. W91CRB-08-C-0135. Additional contributions we made by Curt Olson and Jon S. Berndt as consultants. The resulting parachutist model is unvalidated and therefore, should not be used other than for demonstration purposes. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the US Army RDECOM Acquisition Center.  Due to the significant contributions made to the project by several open source developers,  Creare has released a version of the resulting parachutist model to the open source community for continued development and use.  The FFTK itself continues development as a proprietary Creare project. For information on the FFTK itself, please contact Richard Kaszeta at, or 603-643-3800.
  • Flightgear modeling, animation, and scripting – Curtis L. Olson

Interview: Durk Talsma

Q: What is your forum nickname?

Hehe, guess once. 🙂

Q: How long have you been involved in FlightGear?

Almost since the beginning, actually. I first heard about the project in 1997, when I got an email from Curt Olson, in response to a posting on the usenet newsgroup rec.aviation.simulators.

Q: What are your major interests in FlightGear?

I like the open nature of the project and the possibility to contribute at various levels.

Q: What projects are you working on right now?

I am actually doing several different things for FlightGear. My main project is developing a fully integrated AI air traffic system that contains autonomous vehicles, an ATC system that interacts with both AI controlled aircraft and with the user controlled aircraft. In addition to that, I am one of the editors of the main website, editor of the FlightGear facebook page, involved in the release process, code committer, and organizer of the annual FlightGear booth at FSWeekend in Lelystad (EHLE). In addition, I have recently taken over the administrator role for taxidraw.

Q: What do you plan on doing in the future?

I don’t expect the AI system ever to be finished, so I’m fully concentrating my coding efforts on this project.

Q: Why is it that you are interested in flight simulation or aviation in general?

As a kid I was fascinated by space travel, the Apollo missions to the moon, etc, watching every program on TV, and reading every book I could lay my hands on. As a six-year old, I visited Schiphol (EHAM) airport for the first time, and that sparked my fascination for the big jet airliners. Kind of like every kid at one stage, I wanted to become a pilot. My real interest in aviation didn’t start until I was nearly 20 though, after visiting an airshow at Leeuwarden airforce base (EHLE). This was around the same time as when I got my first PC, a second hand 286DX, which I bought from a relative living in Germany, with a 40 Mb hard disk and 1 Mb of ram. It had a German version of Microsoft Flight Simulator 4.0 preinstalled. So, in addition to learning to “fly” I learned the German word for “crash” as well.

Q: Are you happy with the way the FlightGear project is going?

Yes absolutely. We are currently in the process of further improving our infrastructure, by setting up things like the release plan, formalizing the rules for commit access, aircraft maintenance, and we’re brainstorming about feature requirements for the long term. Firm ideas are present for modularization of the FlightGear code, and some ideas for an integrated launcher GUI have recently been coined in a very informal setting. It will certainly take quite some time before these plans are all realized, but I think that the project is more vital and alive as ever. I’m also just amazed at some of the recent developments, such as Frederic Bouvier’s project Rembrandt, Thorsten Renk’s, local weather system, or Martin Spott’s ongoing efforts to build a unified infrastructure for scenery generation.

Q: What do you enjoy most about contributing to FlightGear?

I think there are a number of aspects that I really enjoy. One of them is the collaboration with other people. Being part of the development team, we’re all pretty much equals, and regardless of one’s age, background, occupation, political or religious conviction, we all share something we like and like to collaborate on. That is really enjoyable. It may also happen that somebody just jumps in and finds a solution in no time for a problem that has been cracking my brain for ages. For example, Adrian Musceac, recent work on generating AI traffic patterns was really something amazing. Likewise, I enjoy the interaction with many other talented people, such as Brett Harrison, who’s just so amazing at making convincing liveries. Obviously there are many other talented people around whom I really enjoy working with and it’s a shame I can’t name them all. Secondly, I also really enjoy having the privilege of being the first to experience a new feature for the first time. I was the first person ever to see the sun and moon appear in a desktop Flight Simulator, and that is a little bit special.

Q: Are there any “hidden features” you have worked on in FlightGear that new users may miss?

Yes, my original contribution to FlightGear was some code to calculate the position of the Sun, Moon, and even the planets. Both the sun and moon are pretty much taken for granted now, but back then (in 1997) FlightGear was the first PC based simulator that actually had a physical rendering of the sun and moon! Nobody will probably even see the planets, but I got the code almost for free, once I figured out how to calculate the solar and lunar positions, so their a little bit of an Easter egg. After finishing the celestial code, and before starting the AI traffic system, I initiated many projects that I subsequently handed over to others. As such, I have extended the time calculation code to deal with local time, and to allow the user control over the time of day, and implemented the original graphical user interface (GUI) system, and the original 2D cloud layers.

Q: What advice can you give to new contributors who want to get started on their first aircraft/new feature/Nasal script?

Be optimistic, be naïve, be realistic, and start modestly. Set yourself an attainable goal! I should probably explain what I mean by this. When we started out, back in 1996-1997, we were what I would now describe as incredibly optimistic in the sense that we believed that we could pull this off, but we were also somewhat naïve in the sense that we really didn’t have any firm idea about the challenges that lay ahead. But, we were able to pull it off, so this shows that we were right after all. But, if you want to contribute don’t start with your magnum opus. Before starting out, take some time to familiarize yourself with the project, get to know the code base, data structure or workflow. In addition, making a good first impression helps. Over the years we’ve seen a tremendous amount of grand ideas and not many of them have materialized, so we’re naturally a little apprehensive you may not find an immediate warm welcome, but if you do come up with a well thought-out idea, you may convince the development team, especially if you can substantiate your ideas with some working code to back it up.

Q: Have you previously used other flight simulators or simulation software in general?

Well, as mentioned before, I started out with FS4, and have pretty much had every version since then, until FS2004. The latter version got me interested in the AI system. When I started playing with the FS2004 equivalent of the ATC system I and began to notice its programming flaws. Determined that I could do this better, I started drawing out my own plans, and since than, I haven’t really touched any other simulator.

Q: How does FlightGear compare in your opinion?

I like FlightGear better because it’s a platform that is constantly moving. I almost exclusively run the cutting edge development version, so occasionally you’re in for a little surprise. Be it positive or negative. But that keeps things a little exciting to me.

Q: Do you remember what first got you interested in FlightGear? How did you learn about FlightGear? In other words, why did you actually download and try FG?

Yeah, that’s a long story. I was reading the usenet rec.aviation.simulators quite frequently at the time, had been exploring Linux for a few years, and finished my C++ programming course at university. This was around 1997, so the Linux distros weren’t as advanced as they are these days, and you still had to do quite a lot yourselves. One particular afternoon, I came across a usenet posting, which read “OPEN LETTER TO ALL FLIGHTSIMULATOR DEVELOPERS”. This was around the time that Microsoft FS97 was the latest version, and many users were dissatisfied. The original poster wanted to write a letter, on behalf of every dissatisfied user, to ask for a better version, asking the big game companies to incorporate their wish list. I responded to the thread, stating that if we really wanted a sim of our own, we should probably do it ourselves. I remember being a little anxious, because I wasn’t sure whether I would actually be able to substantiate that claim, if we were to follow it up. So, a few days later, I was actually a little apprehensive when I opened up my mailbox and found an email from a guy named Curt Olson, inviting me to have a look at, what would eventually become the website. Well, the rest is history I guess…

Q: What was your first impression about FlightGear?

That’s a really interesting question, because there was no FlightGear so to speak of. When I joined, Curt had hacked together a few proof-of-principle demos; the one I downloaded was called linux-demo-0.0.7.tar.gz, if I recall correctly, and it consisted of a small sample of elevation data from a chuck of terrain near Arizona, source code of a primitve (by today’s standards) OpenGL based viewer, a copy of Bruce Jackson’s larcsim FDM, and a simple keyboard interface. But it was exciting to get it to compile, and run!

Q: Compared to other flight simulation software, what are FlightGear’s major benefits in your opinion?

It’s scalability, open architecture, and the fact that it can be a great test bed for ideas, as well as the fact that there is no need for third party add-ons. By bringing every suitable user contribution into a single repository, we essentially create our own add-ons, and in the long run that should remove the burden from the end user to search for extensions.

Q: Do you think it is necessary to know how to program in order to contribute to FlightGear?

No way. In fact it never really has been a requirement, even in the old days when there was a lot more emphasis on C++ development, we already had a need for non-coding developers. Think about documentation writers, etc. These days, the balance is actually really shifting away from programming to artwork. The FlightGear world is essentially still largely an empty place, so we really have a need for high-quality buildings. Many of the exciting developments going on right now are in the development of new scenery textures, 3D modeling, and livery painting. These are actually skills that I essentially lack, so I have a lot of respect for the people working in these areas.

Q: Have you ever used FlightGear professionally or for educational purposes?

I once tried in my previous job, but the computer we bought for the project had serious overheating issues, so the project never really came off the ground. In the mean time, I found a different job, so the project was shelved.

Q: What about FlightGear as a “game”, do you think it can be used as such?

Probably, I like to use FlightGear purely for fun, so usually I just make up my own challenges, such as performing a bad weather landing, taking off and landing on an aircraft carrier, or playing with my latest AI/ATC code. Once finished, the ATC code will add a little bit of a game element, because it will expect you to fly specific routes, arrive at specific locations at a specific time, so as not to clash with other traffic etc etc. The system isn’t finished yet, but with some hacking I did quite recently manage to complete a traffic circuit under guidance of the ATC system, and it’s quite tricky to do right. So, there are some “gamey” aspects of FlightGear that are quite realistic and hopefully challenging. Having said that, I see absolutely no need for any formal gaming rules, or game like features such as setting off explosives and the like. Like many of the other developers, I like to keep FlightGear civil(ized). I don’t object to simulating military aviation though, as long as it doesn’t serve the purpose of glorifying death and destruction.

Q: On average, how much time do you spend working with/contributing to FlightGear?

Hard to say, it varies quite a bit with my day job requirements, but I think on average maybe one or two hours a day.

Q: Which of the more recent FlightGear developments do you consider most interesting/appealing?

There are quite a few. Of the individual projects, I really think that project Rembrandt (Frederic Bouvier’s shadow rendering code) is really exciting. But so is the new effort to unify all the shaders, the atmospheric haze and scattering, and Thorsten Renk’s local weather. I’m also quite happy with the progress we made with the AI traffic/ATC system, even though it’s not finished yet. But, what I think is perhaps even more exciting are some of the long-term infrastructural changes we have recently discussed. I can’t say too much about that yet, because many of the ideas haven’t been formalized yet, but making FlightGear more modularized by making use of HLA technology, and perhaps a more integrated GUI and launcher program are some of the exciting developments that I can see happening in a few years from now.

Q: Is there some feature that you’d truly like to see in FlightGear one day?

Yeah, there are some. Obviously, I’d like to see my own project come to it’s full potential, but in addition to that, I would like to see full scenery development of the polar regions of our planet. One year ago I visited Antarctica in real life, and this is just a very exciting area for flying. I’d also like to see the possibility of lower earth orbital space flight, more seamless terrain textures.

Q: What do you think could be done to attract even more new users and contributors to FlightGear?

Establish a good balance between developing new stuff and doing some public relations work. For the project the key question for survival is not to attract many users, but to attract potential contributors. Obviously, the way to do this is to attract many users, and to hope that there will be a few potential contributors among them.

Q: What about interacting with the FlightGear community? Any tips/experiences you’d like to share?

Nothing really special; just use your everyday courtesy, and keep realizing that we’re all volunteers. I’m usually not that active on the forum or mailing list, but I can tell from 15 years of experience that an intelligent and reasonable response is far more likely to create some momentum than a hurried response that is written in a spur of emotion. Also, I have observed that there is hardly any relation between action and words on either the mailing list or the forum. So when your new to the community, just hang around, get to know the characters and try to establish whom you can trust to be a knowledgeable source of information and who just raises a lot of dust.

Q: Have you ever recommended FlightGear to other users, friends/family?

Not really, my friends and family aren’t really into flight simulation.

Q: Is there anything else you’d like to share with us?

Yeah, have a lot of fun, and if you can try to contribute something to the project.