Tuesday, April 28, 2009

A Modern Cessna

With X-Plane 930 beta 8, we finally integrated the latest version of Max's Cessna 172 SP. Grab the new beta and try it - he's added a number of new features that demonstrate some of the newer X-Plane 9 airplane features.
  • Real 3-d lighting in the 3-d cockpit. Note how the map light tends to illuminate only some of the cockpit as it fades out.
  • 2-d back-lighting on all of the major steam gauges.
  • A bunch of parts can now be dragged in 3-d, including the door handles. This is done via manipulators.
  • Walls! In 3-d cockpit viewer mode you won't be able to leave the airplane until you actually open the doors. The cockpit viewpoint is constrained.
  • The model has the glass parts separated out for correct shadowing, and the glass works correctly from all viewpoints.
  • Panel uses cockpit regions for accurate lighting.
The cirrus jet has been similarly updated. I plan to use the Cessna for a series of tutorials showing how to use these recent X-Plane features.

ATTR_light_level Changed!

I have said this before, but now it's finally true: new file specifications are subject to change in the middle of beta!

In particular ATTR_light_level has changed slightly from beta 7 to beta 8. If you are using this feature in your objects, you will need to update your objects.

A new ac3d beta will be posted later today that supports the updated syntax.

You can read about the syntax here.

Friday, April 24, 2009

So How Big of a Mesh Can You Build?

I get asked about the maximum mesh density in X-Plane a lot.

First, I must note that X-Plane's mesh is adaptive - the triangles are not arranged in a grid, but rather they're arranged to maximize the quality of the mesh with minimal triangle count, while preserving the outlines of water bodies and airports. These pictures show a dynamic mesh fit to a section of the grand canyon (both with and without the shaded DEM for reference):

Second, the actual maximum mesh density is almost certainly limited by system resource constraints. The maximum spatial resolution in a DSF is about 0.6 millionths of a foot. You're going to exceed the maximum number of vertices, or much more likely, the maximum amount of available memory in X-Plane long before you have a grid at that resolution.

(That doesn't meant that resolution isn't usable - remember, the idea of the adaptive mesh is to have very high densities in very improtant places, but not everywhere.)

But in practice this is all moot if you don't code your DSFs by hand - the question that really matters is: what is the highest density mesh that you can use with the current scenery tools.

Until recently, MeshTool was limited to a 1201x1201 (90 meter) input DEM. The input DEM is a simple grid, and the finished DEM is never higher resolution than the input. So until now, it would have been just about impossible to make a really high res mesh, even if X-Plane could handle it.

The most recent changes to MeshTool fix this - MeshTool can now handle a 10801 x 10801 (10 meter) input DEM, which X-Plane should be okay with as well.

This new version will hopefully go into beta relatively soon.

Wednesday, April 22, 2009

ATTR_light_level vs. Generic Instruments

Propsman caught something:
...is modifying the value of a batch of ATTR_light_level tris comparable [performance-wise] with toggling the state of a backlit generic instrument? Instinct tells me that you must have the latter more streamlined than the former, but maybe not?
He is right: in the current implementation, ATTR_light_level is probably a bit more expensive than using generic instruments. This may not be true in the future though.
  • The generic instrument code is pretty tight.
  • Right now ATTR_light_level sometimes has to adjust shaders, which can be expensive.
  • In the future, ATTR_light_level has the potential to be very heavily optimized, while the generic instrument code will always be CPU based.
But to put it in perspective, all instrument drawing is slow compared to scenery drawing - in the scenery world we draw 50,000 triangles of identical OpenGL state in a row, and modern cards do that very, very well.  In the panel, we have to put in a lot of CPU time to figure out how to draw each quad or tri-strip.  Fortunately you probably don't have 50,000 individually programmed flashing lights in your panel.  Heck - there's "only" 3608 datarefs published by the sim.

Perhaps other questions are important when picking ATTR_light_level vs. panel texture:
  • Which is more useful: to be able to have several variant images and variant images that are not "lights" (this is only possible by generics) or the ability to vary the light level gradually and not just have on or off (this is only possible with ATTR_light_level)?
  • Which is simpler to author given the rest of the panel?
In other words, it's all pretty "slow", but fortunately "slow" isn't that slow.  If your light has to blink, you may want to pick what looks best and is straightforward to author.

Datarefs Vs. Commands IV: Duplication

In my previous posts I have tried to explain the difference between commands and datarefs, and when you might use each.  To review:
  • A dataref represents information. You can always read it, and you might be able to change it.
  • A command represents an action.  You can always invoke the action, but you can't tell if it worked without looking at a dataref.
So...why is there so much overlap and duplication?

Dataref Vs. Dataref

There is duplication in the datarefs because we don't delete old datarefs when we add newer, improved ones. The old datarefs stay in place to keep old plugins working. Here are a few reasons why we've added new datarefs:
  • The cockpit2/ and flightmodel2/ sections were added as a new, simpler, easier to use interface for authors in version 9.  (Read more here and here and here.)
  • In some cases, the old dataref was a bit-field while the new one is a simple integer. While plugins can use bitfields, modelers cannot animate using bit fields.
  • In some cases, the old dataref did not represent a clean view of the data. Some old datarefs exposed X-Plane internal structures that are not appropriate for long-term use.
To see this in action, let's look at the autopilot.  How many ways are there to set the autopilot mode?
  1. sim/cockpit/autopilot/heading_mode. This is the original heading mode, and it is marked deprecated, because it exposes a bunch of internal X-Plane autopilot values.
  2. sim/cockpit/autopilot/autopilot_state. This is the ideal autopilot dataref for plugins. It provides all functions, but since it is a bit-field it is not useful for authors.
  3. sim/cockpit2/autopilot/heading_mode. This is a clone of the original heading_mode into the cockpit2 domain. Honestly I am not sure how it got there - I know it was me who put it there, but it sure is a dumb idea; the original dataref is deprecated, so it was stupid of me to duplicate it!
  4. sim/cockpit2/autopilot/heading_state. This is coming in 930 and provides a heading-state enum set appropriate for authors...basically an enum that matches the two heading bits of the autopilot_state dataref that programmers were using.
How do you sort through this? Three rules of thumb:
  • Try to use sim/cockpit2 and sim/flightmodel2 when possible.
  • More recent datarefs are usually better.
  • Use the most useful dataref you can find.
Commands Vs. Commands

Sometimes there is some duplication of commands, e.g.
sim/engines/carb_heat_on                           Carb heat on.
sim/engines/carb_heat_off                          Carb heat off.
sim/engines/carb_heat_toggle                       Carb heat toggle.
Here it's a lot more obvious why there are multiple commands: they affect the carb heat in multiple ways. Typically this is done because commands are mapped to joysticks and other USB hardware; some hardware generates a button press when a command is toggled, but some hardware generates two commands, one for the off and one for the on position.

The rule of thumb is: use the command that gives you the action you want.

Commands Vs. Datarefs

Very often there will be a command and a writable dataref.  Typically we need them both:
  • The command is needed to let users set up their joystick and keyboard.
  • The dataref predates version 9 - writing it was the only way to invoke an action.
Newer datarefs are more likely to be read-only, as we put new "changing the sim" functionality into commands.  To go back to our autopilot example, we have on command: sim/autopilot/heading that lets us arm heading mode.  This command is probably preferable to any of the datarefs for changing the autopilot state.

My previous post discusses writing to a dataref. vs. actuating a command in more detail.

Tuesday, April 21, 2009

Datarefs Vs. Commands III: What Is My command Doing?

It turns out that understanding what a command is doing gets really complicated.

The key idea to untangling it all is this:
  • Commands have duration, and that duration is the amount of time you "hold down" the button or key that actuates the command.
  • Not all commands do things for the entire duration.
An example will clarify this:
  • When you press and hold down the 'p' key to pause the sim, the sim pauses (or unpauses) instantly the moment you press the p key.  Holding the p key down for a long time does not change this.  Pause is a "momentary" command.
  • You have to keep the starter button/key held down for a few seconds to start an engine. If you press it and release it, the starter motor will only run for a fraction of a second, which is not enough time to start an aircraft engine.  Engine start is a "duration" command.
Virtual Datarefs

A "virtual dataref" isn't really a dataref at all - it's a string you can enter into an OBJ or generic instrument that looks like a dataref, but actually is something else.  There is one "set" of virtual datarefs in X-Plane right now.

In X-Plane, you can enter
into a generic instrument or obj animation - this creates a "virtual dataref" around the command's "activation status".  The virtual dataref acts like a read-only integer-type dataref whose value is 0 if the command is not being pressed right now and 1 if it is.

For example, if you animate a push button using CMND=sim/starters/engage_starter_1 (key framing the animation to be "out" when the virtual dataref is 0 and 1 when it is in) then you will have a button that appears to be pressed whenever the starters are engaged.  This will happen no matter how the starter is engaged - your animation will happen whether the user presses a joystick button, holds down a key, clicks on a manipulator, or a plugin runs the command.

Basically, virtual datarefs that provide "activation status" of commands exist so that you can animate the buttons in your virtual cockpit to match what the user is doing with the mouse, keyboard, etc.  These datarefs are read-only; use trigger generic instruments and command manipulators to actually run the command.

For Programmers

In that last section I pointed out that a virtual dataref is not a real dataref at least 3 times. Why am I harping on this?  Well, programmers, you cannot use XPLMFindDataRef to access virtual datarefs yourself.  Sorry.

Virtual datarefs exist to give authors of OBJs and panels access to command activation status, something they would not normally have.

Plugin programmers don't need this - if you are programming a plugin you simply use XPLMRegisterCommandHandler and you will be told when the command is being actuated and released.  Using a command handler is a lot more efficient than reading a dataref because you will receive a call only when the command is pressed, instead of having to check the dataref value every frame.  If you need to monitor a lot of commands, the callbacks are a lot more efficient.

Command Activation Is Not The Same As System Activation

Let's go back to the case of the "pause" command and review.  Here's what we know about the pause command.
  • The command sim/operation/pause_toggle will change the sim's pause state.  The instant this command is pressed, the sim will pause (if it is runnning) or unpause (if it is paused).
  • Holding down the sim/operation/pause_toggle command has no effect beyond its initial press. Hold it down for an hour, it doesn't matter.
  • The virtual dataref CMND=sim/operation/pause_toggle tells you if the pause command is being held down at any instant.
Here's what we know about the pause dataref.
  • The dataref sim/time/paused is 1 if the sim is paused, 0 if it is not.
  • This dataref cannot be written!
So here's what I mean when I say: command activation is not the same as system activation.
  • CMND=sim/operation/pause_toggle tells you if the pause button is being held down.
  • sim/time/paused tells you if the sim is actually paused or not.
  • They are simply not the same!
This becomes important when you start to model the buttons in airplanes.  Take for example an autopilot button for altitude hold.  In many planes, the button can be pushed in, and the moment you do, the altitude hold "arm" light will turn on.  Keep the button held in and it doesn't have any more effect.  Altitude hold arm is a momentary command, but it has a system status indicator light.

I receive a number of questions about how to model this - and the answer is: take advantage of the fact that command activation (is the button pushed in) and system activation (is the light on) are not the same.
  • You would use a read-only dataref for the autopilot state to turn on the indicator light.  (There are two ways to do this: use ATTR_light_level to turn a _LIT texture on and off, or use panel texture and map a generic instrument like a rotary or an annunciator.)
  • You would use a command to make the button work.  Probably you'd use a command manipulator on the button mesh.
  • You would key frame the button's animation based on the virtual dataref wrapped around the command you are using with the manipulator.
The result will be a button that lights up when the AP is armed but pushes in while it is being pressed (whether via the mouse in the 3-d cockpit or a joystick button press or a keyboard button press).

In my final post, I'll comment on how much overlap there is between datarefs and commands.

Monday, April 20, 2009

Datarefs Vs. Commands II: Which One Should I Use?

There is a lot of overlap between the datarefs and commands; very often there is both a dataref (telling information about some part of the sim) and a command (which takes action to change some part of the sim).  Which should you use?

Here are some general guidelines:
  • If you want to set up a joystick or keyboard, you have to use a command. The joystick and keyboard configuration dialog box lets you associate actions with a keystroke or button press, not information!
  • If you need to show the status of a system (E.g. "is the landing gear down") use a dataref. I will cover this issue in more detail in part 3, but basically only datarefs show you information.
The ambiguous case is whether to use a dataref write or a command to change a system when both exist.
  • If there is a command that exactly does what you want to do, prefer the command over the dataref.  For example, it is better to arm the autopilot using the commands than the datarefs.  Changing the autopilot state often involves changing a lot of variables at once in complex ways.  When you issue the command, that work is done for you, correctly, every time.
  • If the command is not really suitable for your purpose, use a dataref.  For example, to change the engine throttle position, do not use the command sim/engines/throttle_up to move it up "a little bit."  Use the dataref sim/cockpit2/engine/actuators/throttle_ratio to set the throttle to the precise position you want.  The throttle-up command exists so that users with no joystick or mouse wheel can fly with the keyboard by pressing the F1-F2 keys (bound to throttle-up, throttle-down).  It is not meant to precisely control the throttle position!
(I will discuss why there are so much overlap between commands and datarefs in part 4.)

DataRefs Vs. Commands I: What's The Difference

What is the difference between a dataref and a command? They serve different purposes in X-Plane, but it's easy to get them confused, especially because the names can look so similar. If you only take one thing away from this comparison, it should be:
  • Datarefs are information.
  • Commands are actions.

A dataref is a single bit of published information. For example, the user's indicated airspeed, as seen by the pilot, is a dataref, stored in:
Datarefs have names that do not change. Datarefs made available by X-Plane start with sim/ while datarefs made available by plugins start with another prefix. Datarefs have been in X-Plane since the release of the plugin system in version 6.70.

You can always read a dataref, but sometimes you can change it. Trying to change a dataref usually has one of three actions:
  • If the dataref is not writable at all, nothing happens.
  • If the dataref is writable, it will change.
  • Sometimes a dataref may be writable, but only after changing some other sim configuration. For example, you can only "write" to the control surface deflection datarefs after setting the control surface override dataref to 1. (If you don't set this override, X-Plane will constantly write its own ideas of the control surface positions to the control surface datarefs and your changes will be lost.)
You can read and write datarefs:

A command is an action that the sim can take on your behalf. For example, the command
arms the autopilot for altitude hold.

Like datarefs, commands have permanent names, starting with sim/ for X-Plane or other prefixes for plugins. Commands have been available in X-Plane since version 9.0.

You can always actuate a command, but there is no guarantee that it will do anything. For example, the engine starter command won't start the engine if the plane has electrical starters and the battery is dead.

You can use commands by:

Plugins can add both new datarefs and new commands to the sim. Plugins can also change the behavior of all built-in sim commands, and can change the information in some datarefs.

Where Do I Find Datarefs And Commands

X-Plane's default commands and datarefs are listed in the text files Commands.txt and Datarefs.txt in the Resources/plugins folder. (Note: providing the command list is new to X-Plane 930.) The dataref list is also available on the X-Plane SDK Wiki.

Up next: when should I use a command and when should I use a dataref?

Friday, April 03, 2009

The X-Desktop and iPhone Are Not Zero-Sum Games

On the menu this morning: first a whiny rant, then a nerdy one.

Someone pointed me at this post. Before I go into my geeky diatribe about leveraging code, a quick note: it is true that at this point Austin is working heavily on the iPhone - probably more on the iPhone than the desktop. It is also true (but not mentioned) that Laminar is investing more man power into X-Plane for the desktop now than it ever has in the past. (With the iPhone we've grown our workload with a second "front" for products, but we've increased staffing a little bit too.)

It is also true that X-Plane 9 free updates have been less frequent. This doesn't mean there's less code going into them - it just means that we're doing 4-5 months of coding and 3 months of beta instead of 2 months of codindg and 1 month of beta. I'm not sure if this is better or why it is happening (each release has to be individually planned for the circumstances) but one observation:

Given how many video cards and drivers are out there and how they react differently to the X-Plane code, I wouldn't want a beta process less than 2-3 months, because I want to know that it's had time to run on a wide variety of hardware. If our beta has to be at least 2-3 months, then a 3-month release cycle would have us in beta all the time!

My whiny rant (the short version) is this: Laminar Research isn't in a position (as a company selling a product) to post all of the details of how our business operates internally. So posts that say "Laminar should do X" (where X is a business decision) drive me a bit nuts, because I can't post a reasonable reply.

Now here's my real point: development for the desktop and the iPhone are not a zero sum game. Clearly we leveraged the desktop sim to create the iPhone app. But it goes both ways; work on the iPhone also benefits the desktop.

Here is one example: when beta 8 comes out, Plane-Maker's "export to object" will a much more complete OBJ, in OBJ8 format, with animations already in place for things like wing controls surfaces!

That's a feature that people have been asking about for a while - both to make CSL objects* and as a way to save time when moving from a Plane-Maker drawn plane toa custom one. (You would first export an animated OBJ8, then start adding details in a 3-d modeling program. With beta 8, you won't have to rebuild your control surfaces from scratch.)

But...it's also a feature that is critical to the iPhone! The iPhone version of X-Plane uses animated OBJ8 files as well; this feature helped us internally to prepare planes for the iPhone version, but it's also a requested feature from our authors.

So my response to those who say that the iPhone has taken away from X-Plane development is that it is not true - not only because we are developing more heavily for the desktop, but because sometimes iPhone development is for the desktop.

* I do not remember whether the current compiled versions of libXplaneMP use OBJ7 or OBJ8 objects. Since libXplaneMP uses the OBJ code from XPTools, it should be (relatively) straight forward to update libXplaneMP and the clients that use it to use OBJ8 directly. In the meantime, you can use ObjConverter to convert an exported OBJ8 back to OBJ7 for CSL use.

Wednesday, April 01, 2009

Scenery Tools Progress? Yes.

I've been hacking at the scenery tools over the last few days...and I think we're making some progress.  I mention this because it can be hard to see the work getting done - there is nothing for months, then a new version of a tool.  (Well, now you can watch the code change as it happens.  If you don't think this like watching paint dry, you might be a geek!)  I also mention it because two of the biggest improvements aren't my own - they're insanely helpful user contributions.

The AC3D Plugin

Version 3.2 is now in beta.  Version 3.2 catches the plugin up with all of the latest OBJ tricks, including dataref-controlled LIT levels, manipulators, hard surfaces, etc.  This might be more useful for airplane modelers than scenery modelers, but I think the variable _LIT channels can be used for some interesting scenery effects too.


WED 1.1 will go beta once X-Plane 930 is out of beta, and will feature (very primitive) overlay editing.  It's not a thing of beauty (yet) - the overlay editing features are leaner than the airport editing tools were.  But 1.1 will allow you to edit an overlay and airport in one workspace.

I can tell you now that the area where WED 1.1 is weak is preview.  But I don't want to hold off the ability to edit overlays in WED "at all" just because it's not as nice as it could be (and someday will be).

It should be possible to work on an overlay project in both WED and OverlayEditor - for example, you could build a taxiway layout in WED, using the overlay capabilities for custom pavement and lines, but use OverlayEditor to place 3-d objects (with a 3-d view).


To be blunt, MeshTool in its current form is really hard to use.  Part of that is expected - MeshTool is a "back-end" tool like DSF2Text, aimed to simplify the problem of creating base a base mesh for some other program.  But even as a back-end, MeshTool is difficult to use because it has bugs in its polygon processing, which often result in cryptic and hard-to-diagnose crashes when used with even relatively simple input data.

The good news is that I have my base mesh generation code running with a new polygon processing library (CGAL) - the switchover from my old buggy code to CGAL was done by Andrew and contributed back as a patch.  I've been throwing all sorts of weird data at the polygon cutting algorithms and they are rock solid.

The mesh generation code is part of MeshTool, so the next build of MeshTool will feature these processing improvements.  That's where Janos' extensive work on the build environment comes in - he's basically turned the scenery code from a big pile of my mad ramblings into a sane multi-platform project.  This will mean a much simpler, faster process for releasing tools, particularly on Windows and Linux.

(Previously once I got a tool working, I would have to waste a few days trying to get the code to work on Windows too, slowing the whole process down.)

I expect to recut the "scenery tool" pack (that is, MeshTool, DSF2Text, and all of those other smaller apps) some time during WED beta, after 930 goes final.  If all goes well, there should be native versions of the command-line tools for Linux.

MeshTool Needs a UI

Fixing MeshTool's polygon processing is necessary, but not at all sufficient to make it a useful tool; to be really useful, it needs a real user interface.  I think there are two strong possibilities for this:
  1. WED could develop base mesh editing.  Under this scheme, you would be able to draw land use polygons in WED and create a "DEM" object for the elevation.
  2. MeshTool could be an output option for QGIS (via some kind of plugin or script).
I don't know which it will be yet.  The advantage of WED is that an author could have all scenery elements in one place.   The advantage of QGIS is that it contains advanced GIS features that I will probably never put in WED.  It may be that QGIS's UI is simply not intuitive to less advanced users - I don't have good perspective on this.