Re: [Flightgear-devel] Possible new thinking for 2D/3D cockpit instruments
On Thu, 2005-12-01 at 22:47, Josh Babcock wrote: Steve Hosgood wrote: Also, have you considered looking into OpenGC? It won't give you the MSFS like functionality of dragable sub windows, but I think it would allow you to make arbitrary windows to display instruments in cutouts. I was deliberately thinking that you **don't** want to use OpenGL for that sort of thing. The GPU has enough work to do rendering the view out of the windows, it would be a waste of its time rendering instruments for the fascia - they're always going to be displayed straight on with flat lighting. It's just a simple animation job for a normal window. I see that a cockpit building discussion has kicked off in a parallel part of this thread... Steve ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Possible new thinking for 2D/3D cockpit instruments
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Steve Hosgood schrieb: I was deliberately thinking that you **don't** want to use OpenGL for that sort of thing. The GPU has enough work to do rendering the view out of the windows, it would be a waste of its time rendering instruments for the fascia - they're always going to be displayed straight on with flat lighting. It's just a simple animation job for a normal window. That's why the future of the 2D desktops will be rendered by the 3D hardware (Windows Vista, the OpenGL based X-Server, ...). A while ago 2D desktops would profit from the graphic accelerator graphic cards. They had chips that could draw very fast lines, etc. pp. But today we've got 3D accelerators that can do even more. They are even programmable. So the new OSes use that functionality for a fast visual feedback. So it doesn't make sense to pass the rendering of some instruments back to the OS. It will just give it back to the graphics adapter - with the aditional overhead of going through the OS. The only alternative to reduce the load on the GPU is to draw it with the CPU by hand (note: this is really CPU intensive!). But if the CPU idles too long (what I really doubt) we could easily increase the FDM resolution, AI traffic, ... CU, Christian -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.0 (MingW32) iD8DBQFDkC5JlhWtxOxWNFcRArQ2AJ0Y9W2z2ZlrQ3615T3LVUGOv3T10QCgq1Ac Lv9HbthiUs1IqdPu6uq5ZNo= =rjDA -END PGP SIGNATURE- ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Possible new thinking for 2D/3D cockpit instruments
On Thu, 2005-12-01 at 20:23, Curtis L. Olson wrote: This is why all those oddball home/hobby cockpit builders aren't as far off their rockers as it might first appear. They are taking a huge step towards a more realistic simulation environment. Dead right. I'd never knock them - more like admire their enthusiasm. And as you say, an open source sim like FlightGear is much more likely to be able to interface to all these home-made instruments. And I'm sure all these people have spouses who understand the importance of a realistic flying experience. Yeah, right! :-) having the flight model right exactly on, is less important than having a full scale cockpit with controls that have the right amount of force feedback at the right times. An enclosure is a huge addition... Not sure about whether FlightGear currently allows for force feedback, but of course, if anyone's flight sim could do it, FlightGear could. Does FlightGear provide output data that would allow you to tip a cockpit on hydraulic rams (or any other system) to try and model changing G forces for the pilot? I've mentioned the possibility in other mails on this thread for a revamped future FlightGear instrument model to cater for separate windows (maybe on other display heads of course) which would help implement fascias using LCD panels behind cutouts. I'd have thought that the cockpit-builder types would be clamouring for such an addition, yet no-one's apparently all that enthusiastic. Do the current crop of cockpit builders happen to use real simulated physical instruments wired to USB or something? I read elsewhere that the 747 guys were simulating a glass cockpit, so maybe they didn't have any physical instrument scenarios to cope with. Hasn't anyone tried a cockpit-build for a WWII plane with FlightGear yet? Steve. BTW, nearly unrelated - one of the Discovery channels in the UK recently ran a documentary on recreating the Dambusters raid on the Ruhr in 1943. They had a (rather crude looking) mockup of a Lancaster bomber and a crew of modern RAF types who tried to simulate reproducing the raid. Whose flightsim was that? Unlikely to be FlightGear, unless the TV people commissioned their own Lancaster FDM. Did anyone apart from me see it? It looked like the instruments panel for the Lancaster was simulated with the old lcd panel behind holes cut in plywood trick. Actually, I'm not even sure they bothered with the plywood. They certainly didn't appear to bother with putting a skin on the fuselage of the fake plane - they just ran it in a darkened warehouse. ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Possible new thinking for 2D/3D cockpit instruments
Steve Hosgood wrote: On Thu, 2005-12-01 at 22:47, Josh Babcock wrote: Steve Hosgood wrote: Also, have you considered looking into OpenGC? It won't give you the MSFS like functionality of dragable sub windows, but I think it would allow you to make arbitrary windows to display instruments in cutouts. I was deliberately thinking that you **don't** want to use OpenGL for that sort of thing. The GPU has enough work to do rendering the view out of the windows, it would be a waste of its time rendering instruments for the fascia - they're always going to be displayed straight on with flat lighting. It's just a simple animation job for a normal window. I see that a cockpit building discussion has kicked off in a parallel part of this thread... Steve ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d No, OpenGC ^ http://www.opengc.org/ Josh ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Possible new thinking for 2D/3D cockpit instruments
On Fri, 2005-12-02 at 14:28, Josh Babcock wrote: No, OpenGC ^ http://www.opengc.org/ Oops. Sorry. Steve ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Possible new thinking for 2D/3D cockpit instruments
Steve Hosgood writes: Do the current crop of cockpit builders happen to use real simulated physical instruments wired to USB or something? There are several vendors out there who have simulated instruments with needles and the like, often driven by RC servos. Granted, this runs your price up $20 an axis, probably on the order of $65-100 per instrument if you're not good enough with the styrene to build them yourself, but if you think about how much LCD panel space it'd take to cover a real-sized keyboard, I think it starts to look like a reasonable trade-off. Especially since external instruments can be driven with almost zero main CPU. if I had the budget to go hardcore (I've only built pedals, a full-length stick and a collective lever, but I've been looking at the USB joystick spec 'cause I'm starting to think about 10 bits and the ability to do some stuff on microcontrollers) I'd want that extra LCD to be dedicated to view, not instruments. Admittedly, though, I'm interested in aircraft that have a limited instrument set. Folks interested in comercial airliner cockpits have it harder than those of us who are into helicopters that just because of stability and the difficulty of removing hands from the primary controls to diddle with knobs would need a copilot to do any IFR. Dan ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Possible new thinking for 2D/3D cockpit instruments
--- Steve Hosgood wrote: snip I propose then that every single instrument on the cockpit has the ability to be double-clicked, and if so then a separate draggable window appears containing a magnified view of that same instrument. Hi Steve, Personally I think this is a fine idea, and indeed gets around the challenge of having generic dialog boxes that don't match specific instruments. I'm not sure whether it would need to be magnified, it would be sufficient just to display them at normal resolution (e.g. 128X128 for most round gauges, 256x60 (ish) for radio stacks). However, I don't know whether we can easily display the gauges in windows by themselves - I'm not familiar with the graphics routines we have, but I suspect that we are stuck with a single rendered window (as opposed to dialogs created using GUI widgets), unless we want to take huge perf hits. I think the main issue here is with the radios, GPS and autopilot. One simple solution would be to create a radio panel for the plane containing these components, the visibility of which could be toggled either from the menu, or from a keypress. The downside to this approach are: - each plane would need to have the panel created specifically (unless someone wants to write a generic routine to pick up all the appropriate controls and automagically generate a panel on the fly) - The panel couldn't be dragged and dropped - though it could be shifted using the normal controls. - I don't think we'd be able to use the double-click idea, as that normally causes two increments/decrements. On the plus side: - I think this is quite easy to implement using the existing code, so is fairly safe for v1.0.0 - Just about anyone could do it - it's just a bit of messing around with panels, so we can all pitch in. - You can bind the key normally used for the radio dialog to displaying this specific panel. -Stuart ___ Yahoo! Messenger - NEW crystal clear PC to PC calling worldwide with voicemail http://uk.messenger.yahoo.com ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Possible new thinking for 2D/3D cockpit instruments
Flightgear (and any other flight sim) is trying to reproduce the experience of flying, both in terms of the flight dynamics and (to a limited extent) the whole experience. As such, many of the instruments in the virtual cockpit can be configured with mouse-clicks on the instruments themselves. Some can also be configured through dialog boxes. If FG wants to try and model the flight experience, these alternative dialog-box UIs must go. There are no pull-down menus on a real plane, and no dialog-boxes. Providing them therefore breaks the flight experience. I disagree with the fact that it breaks the flight experience. On the real plane, you extend your hand and twist a knob. Unless you're building an external hardware to augment your flight simulation experience (i.e., actual radio stack panel with the knobs to twist that will interface the PC), you will not have the same experience. Touchscreen might be smth, but most of us don't have a touchscreen either. Doing a mouse click on a radio knob (that is rendered to a tiny circle less than the natural size as the whole screen is less than 1:1 at the default zoom where you see both the window and the radios) is thus significantly more difficult and a more time consuming task. BTW, I am comparing it to real flight experiences. Mostly you even twist these knobs BLIND in the real life, only occasionally glancing at the frequency displays when you make the approximately correct amount of clicks, and look outside. No way to model that w/o a real knob. There is a concept of flow in real flying, referring to the flow of your hands around the cockpit, and the only way to train these is to do it in 1:1 scale 3D physical environment. Clicking will not give you the correct flows, because your hand doesn't move the same. Therefore, a way to do it via keyboard shortcuts/dialogs is a reasonable compromise --- you want to be able to make it with an approximately same ease. If, however, you want to do the clicking, that's all right, too --- but please back off from the idea that everything but the clicking must go. Vassilii ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Possible new thinking for 2D/3D cockpit instruments
I just struck me that it's already possible to get a better look at the instruments, both 2D and 3D, in a very simple way: I think all OS's and windowmanagers have a magnifier tool. It can't magnify beyond the screen resolution of course (640x480 would still be 640x480), but it solves the problem with blurred tiny characters on small weathered monitors, like mine. ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Possible new thinking for 2D/3D cockpit instruments
Steve Hosgood wrote: Makes me wonder whether there's an excuse for some new thinking on the subject of UI design, regardless of whether a cockpit is 3D or 2D. Here's what I propose - please be kind with your comments, I'm not trying to dictate terms or tread on anyone's toes: I propose then that every single instrument on the cockpit has the ability to be double-clicked, and if so then a separate draggable window appears containing a magnified view of that same instrument. Obviously, it will be a *lot* easier to click on buttons and knobs on this magnified instrument, though some people with colossal screens won't need to bother and can carry on with the normal-size instruments. Just as a note, this functionality already exists. You can use the mouse to look around and zoom in. Zoom in, click, zoom out. I do it all the time. Just my $0.04 Now just off to don fireproof suit Heh, I'll try to keep the temperature down :) My personal view is that clicking on a little box on the screen is nothing at all like reaching out and touching/feeling a knob or switch. I don't think that the mouse can come anywhere close to providing a real experience, nor is it capable of even supplying the kind of effectiveness that a ergonomically designed (real) cockpit can provide. This is not to say that I am against making everything in the cockpit clickable, I think it's great that the functionality is there and I try to provide it when I am designing a cockpit, but I also recognize that there are a lot of people out there (myself included) that would much rather use the keyboard, dialog box or pulldown menu. In short, It's all well and good to add a functionality, but talking about taking away a functionality that someone wanted enough to go to the trouble of creating is not productive. If you don't like it, don't use it. If you want something that doesn't already exist, add it. Josh ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Possible new thinking for 2D/3D cockpit instruments
On Thu, 2005-12-01 at 11:56, Buchanan, Stuart wrote: I propose then that every single instrument on the cockpit has the ability to be double-clicked, and if so then a separate draggable window appears containing a magnified view of that same instrument. Personally I think this is a fine idea, and indeed gets around the challenge of having generic dialog boxes that don't match specific instruments. I'm not sure whether it would need to be magnified, it would be sufficient just to display them at normal resolution (e.g. 128X128 for most round gauges, 256x60 (ish) for radio stacks). On my screen I can just about read the round gauges and radios and can't read the compass (on the 3D Cessna). The 2D cockpit is much more legible. Hence the comment about magnification. However, I don't know whether we can easily display the gauges in windows by themselves - I'm not familiar with the graphics routines we have, but I suspect that we are stuck with a single rendered window (as opposed to dialogs created using GUI widgets), unless we want to take huge perf hits. I was thinking of using the machine's underlying windows system for these popup per-instrument displays. So that would be GTK or similar on X-Windows and native MS Windows API on M$Windows machines. Alternatively, FlightGear could standardise on GTK for writing the things, and use the fact that there already are GTK-on-M$Windows and GTK-on-MacOS libraries out there that would take care of the platform-dependencies. I think the main issue here is with the radios, GPS and autopilot. One simple solution would be to create a radio panel for the plane containing these components, the visibility of which could be toggled either from the menu, or from a keypress. I would leave the OpenGL engine to display the panel as it does now, but some instruments (at the users' discretion) may get duplicated in their draggable, scalable windows. - The panel couldn't be dragged and dropped - though it could be shifted using the normal controls. It could, if it was written to employ the ordinary windows-system windows, not be part of the OpenGL main display window. - I don't think we'd be able to use the double-click idea, as that normally causes two increments/decrements. Double-click is normally detected as such really low down and doesn't normally get confused with two single-clicks. - You can bind the key normally used for the radio dialog to displaying this specific panel. Indeed. And there's another possible plus side. There was a thread here a few weeks back about the serious flightsim-heads who like to build physical cockpits and have real instruments. Apparently, one way to get the effect of real instruments on a budget is to fit an LCD panel behind cutouts in a fascia plate and display the instruments on that LCD panel. Well, doing this gets a load easier if we've already written the code for every instrument to be able to render itself (magnified) in photo-realistic style in individual windows. The builder of a fake cockpit can then drag all the magnified instruments to wherever they're needed behind the cutouts in the fascia, and hey presto! Job done (pretty much). FlightGear would need to be able to remember at start-up how the user wants to display any instruments that have to start in this windowed mode: i.e their magnification (or window-size) and window location (X and Y coords on which physical screen). Steve. ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Possible new thinking for 2D/3D cockpit instruments
On Thu, 2005-12-01 at 14:15, Josh Babcock wrote: Just as a note, this functionality already exists. You can use the mouse to look around and zoom in. Zoom in, click, zoom out. I do it all the time. That's a very good trick (just tried it). Never thought of that one, and yes, I can even read the buttons on the autopilot by doing that. The only trouble with that approach is that you can't both look out of the window *and* read the autopilot without quite a few mouse-clicks and some x/X keypresses. I'll grant you that it does allow you read small buttons and things, it's a great workaround for my main gripe. I also recognize that there are a lot of people out there (myself included) that would much rather use the keyboard, dialog box or pulldown menu. In short, It's all well and good to add a functionality, but talking about taking away a functionality that someone wanted enough to go to the trouble of creating is not productive. Yeah, OK. Several people have said the same thing now, so obviously the dialog-box option is regarded as a must-have. As you say, let's not throw out something that works. However, can the implementation be changed so that repeats of the autopilot snafu can't happen? I suggested in a different reply that maybe the instrument object should be in charge of all its related displays - whether that's the OpenGL one, the dialog box or (if it was accepted as a good idea) my separate pop-up window alternate-view idea. If you don't like it, don't use it. I get the message! (And I have no problems with that.) Steve. ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Possible new thinking for 2D/3D cockpit instruments
windowmanagers have a magnifier tool. It can't magnify beyond the screen resolution of course (640x480 would still be 640x480), but it solves the problem with blurred tiny characters on small weathered monitors, like is it not the same effect as if the characters are rendered w/o antialiasing? Is it possible to do from within flightgear (to render them in this way)? ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Possible new thinking for 2D/3D cockpit instruments
Ralf Gerlich wrote: Heh, I'd like to see you looking at the Autopilot _and_ out of the window in a real plane. ;-) As was mentioned, the nearest you could come to the flow in the cockpit IRL - not looking at the instrument and still changing its setting - is probably using the keyboard...at least as far as I see that as a pure simulation pilot ;-) This touches on one of the *big* differences between a 'toy' and a real pilot training tool. Having the entire instrument panel available at it's correct scale and location as well as having all the cockpit controls in their right location with the right amount of force feedback is a *huge* thing in terms of making the simulator realistic. This is why all those oddball home/hobby cockpit builders aren't as far off their rockers as it might first appear. They are taking a huge step towards a more realistic simulation environment. And I'm sure all these people have spouses who understand the importance of a realistic flying experience. You could have *perfect* flight dynamics that nailed all the numbers and all the nuances of the model exactly right, but if you are sitting at your desk, holding a $20 joystick in one hand and typing on your keyboard with another, while peering at a 17 monitor ... it's just not going to ever be all that realistic of an 'experience.' I will even go so far as to assert that when creating a 'realistic' flying experience, having the flight model right exactly on, is less important than having a full scale cockpit with controls that have the right amount of force feedback at the right times. An enclosure is a huge addition because it blocks out many of the real world distractions that can snap you back to reality. In addition, assembling a wrap around visual system that projects a field of view that exacatly matches the field of view covered by your display device is also very helpful. All of this is said from the perspective of creating a realistic flying experience. If you are using flightgear for other purposes (such as an engineering simulator or visualization tool, running it on a desktop PC or laptop may be exactly what is needed.) Regards, Curt. -- Curtis Olsonhttp://www.flightgear.org/~curt HumanFIRST Program http://www.humanfirst.umn.edu/ FlightGear Project http://www.flightgear.org Unique text:2f585eeea02e2c79d7b1d8c4963bae2d ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Possible new thinking for 2D/3D cockpit instruments
Curtis L. Olson wrote: Ralf Gerlich wrote: Heh, I'd like to see you looking at the Autopilot _and_ out of the window in a real plane. ;-) As was mentioned, the nearest you could come to the flow in the cockpit IRL - not looking at the instrument and still changing its setting - is probably using the keyboard...at least as far as I see that as a pure simulation pilot ;-) You could have *perfect* flight dynamics that nailed all the numbers and all the nuances of the model exactly right, but if you are sitting at your desk, holding a $20 joystick in one hand and typing on your keyboard with another, while peering at a 17 monitor ... it's just not going to ever be all that realistic of an 'experience.' One of the knocks from the May show ( which is totally my fault) was the cheezy joystick. So here we were with a full scale 747 glass cockpit with a large screen plasma OTW display running top of the line flight dynamics (JSBSim), world class scenery (FlightGear), high fidelity subsystem models (Mathworks), and a noodle for control. If we had had a decent control system, we could have faked the rest ;-) JW ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Possible new thinking for 2D/3D cockpit instruments
John Wojnaroski wrote: One of the knocks from the May show ( which is totally my fault) was the cheezy joystick. So here we were with a full scale 747 glass cockpit with a large screen plasma OTW display running top of the line flight dynamics (JSBSim), world class scenery (FlightGear), high fidelity subsystem models (Mathworks), and a noodle for control. If we had had a decent control system, we could have faked the rest ;-) Come on Jack, when are you going to drive up to Mojave with your hacksaw? I can get you past the fence, the rest is up to you. :-) Curt. -- Curtis Olsonhttp://www.flightgear.org/~curt HumanFIRST Program http://www.humanfirst.umn.edu/ FlightGear Project http://www.flightgear.org Unique text:2f585eeea02e2c79d7b1d8c4963bae2d ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
Re: [Flightgear-devel] Possible new thinking for 2D/3D cockpit instruments
Curtis L. Olson wrote: John Wojnaroski wrote: One of the knocks from the May show ( which is totally my fault) was the cheezy joystick. So here we were with a full scale 747 glass cockpit with a large screen plasma OTW display running top of the line flight dynamics (JSBSim), world class scenery (FlightGear), high fidelity subsystem models (Mathworks), and a noodle for control. If we had had a decent control system, we could have faked the rest ;-) Come on Jack, when are you going to drive up to Mojave with your hacksaw? I can get you past the fence, the rest is up to you. :-) I hear you. There is also a boneyard at El Mirage which has some hulks that go back to WWII. I understand the tv series LOST got some of the props from that site. I ought to give Tom a call now that the daytime temps up there have become bearable. Just a question of time and energy. The design issue is how to keep it portable so we can haul the gear around to shows like Scale4x coming up in Feb 06. Same problem with putting everything into a shell, fantastic for a fixed installation but kind of like the old story of the fellow who builds the 30 foot sailboat in his cellar Regards John W. ___ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
