FlightGear Qt launcher

This article or section contains out-of-date information Please help improve this article by updating it. There may be additional information on the talk page.

FlightGear Qt launcher

The aircraft page of the Qt launcher for FlightGear 3.5 as rendered on Windows 7
Started in	01/2015
Description	Integrated launcher for FlightGear designed to replace FGRun
Maintainer(s)	James Turner
Contributor(s)	James Turner
Status	Under active development as of 05/2016.
Folders	flightgear/flightgear/next/src/GUI
Changelog	flightgear/flightgear/next/src/GUI log view

The Qt launcher is an integrated launcher for FlightGear designed as a replacement for FGRun. Initially designed as a stop-gap solution in FlightGear 3.4 for the problem of support for the old FlightGear Mac launcher ending with OS X Yosemite, it has become FlightGear's main launcher for all platforms, and is shipped with all official FlightGear releases (however, users building from source may need to separately configure/rebuild FlightGear).

History and status

In October 2014, Apple released OS X 10.10 Yosemite. Unfortunately, one of the frameworks the old FlightGear Mac launcher relied upon, called RubyCocoa , was removed, making it incompatible with OS X Yosemite. James Turner started work on a solution for the then-upcoming 3.4 release. He added a simple launcher using Qt, run before the main window would be created. It was released with FlightGear 3.4 as a Mac-only feature.

After FlightGear 3.4, it was decided to that this temporary Mac-only launcher would be developed into a replacement for FGRun and that it would become part of the plan for updating the FlightGear user interface. In FlightGear 3.6, it became available for all platforms, and has continued to be developed, enhanced, and refined.

As of May 2016, the Qt launcher is under active development. This includes the adding of new features, fixing bugs, and refining existing features.

Usage

To run the launcher:

• Microsoft Windows users can double-click the FlightGear <version> icon on the Desktop or click on Start -> All Programs -> FlightGear <version> -> FlightGear Launcher;
• OS X and Linux users must open a terminal and run fgfs with the option --launcher:

$ fgfs --launcher

Note  The launcher is an optional feature for the reasons explained below in the Dependencies section. As a result, if you are on Linux and have installed the version of FlightGear provided by your distribution, it might not be available because the packagers might have not enabled it. In that case, ask your distribution to enable the feature.

• 

  Location page of Qt launcher in FG 3.5 on Windows 7

• 

  Settings page of Qt launcher in FG 3.5 on Windows 7

• 

  Addons page of Qt launcher in FG 3.5 on Windows 7

Multiplayer connections

As the launcher is still under development, only some basic options are present; notably, no multiplayer option is shown. To connect to the multiplayer servers, either:

• start FlightGear from the launcher and then choose Multiplayer -> Multiplayer Settings inside the simulator;
• or open the Settings tab in the launcher and specify the following parameters:

--console --multiplay=out,10,mpserver01.flightgear.org, --multiplay=in,10,, --callsign=mycallsign

replacing mpserver01.flightgear.org with the multiplayer server to be used and mycallsign with the desired callsign.

Enabling console display on Windows systems

To enable console display on Windows systems, right-click on the FlightGear shortcut on the Desktop and click Properties. In the Target box, add --launcher at the end and click OK.

Aircraft/TerraSync download directory

As of FlightGear 2016.2, the location used by the launcher to store downloaded aircraft and TerraSync scenery can be specified in the Add-ons tab. Specifying the TerraSync download directory manually has, thus, become unnecessary.

Background

In accordance with the FlightGear 4.xx Roadmap, FlightGear's current graphical interface (based on the PUI library of PLIB) will be replaced by one based on Qt. FGRun will also be removed in future to be replaced by this launcher. In addition, a browser-based UI (called Phi) will continue to be developed in parallel. This will result in the following user interface:

• Qt launcher
• Internal UI based on Qt, possibly with Canvas integration
• External browser-based UI (Phi)

However, it should be made clear that Qt will always remain an optional dependency.

Preferences

In FlightGear 2016.2 and later, the preferences are stored in $FG_HOME/flightgear.org/FlightGear.ini. Older versions make use of the default location used by Qt:

• the Registry on Microsoft Windows;
• ~/Library/Preferences on OS X;
• the default preferences location on Linux (usually ~/.config/FlightGear/FlightGear.conf).

Preferences can be reset to their default values by holding down the Alt key while starting the launcher.

Repercussions

Dependencies

When the launcher was introduced, there were significant divergences of opinion regarding whether it was appropriate to require the dependency on Qt; the core developers concluded that it was undesirable, unless it had a specific benefit. Since, for the time being, Qt is only needed for the launcher, it has been made an optional dependency, and people can continue using other frontends should they wish so.

If CMake is not able to find Qt 5.1 (the minimum required version) or later, the launcher is not compiled, and the following message will be printed:

-- Qt launcher enabled, checking for Qt 5.1 / qmake
CMake Warning at CMakeLists.txt:303 (find_package):
  By not providing "FindQt5.cmake" in CMAKE_MODULE_PATH this project has
  asked CMake to find a package configuration file provided by "Qt5", but
  CMake did not find one.

  Could not find a package configuration file provided by "Qt5" (requested
  version 5.1) with any of the following names:

    Qt5Config.cmake
    qt5-config.cmake

  Add the installation prefix of "Qt5" to CMAKE_PREFIX_PATH or set "Qt5_DIR"
  to a directory containing one of the above files.  If "Qt5" provides a
  separate development package or SDK, be sure it has been installed.

As of May 2015, the Qt launcher is the default on Microsoft Windows and OS X: in the case of Linux, it's up to the distribution packagers whether to include it or not.

If, how and when the internal GUI can leverage Qt is under investigation and in such an early state that those responsible preferred to not announce anything yet. If this ever happens, this will be discussed on the mailing list.

Segmentation faults and race conditions

Note  Also see ticket #1819

Should installing an aircraft to the proper aircraft folder be allowed to crash the launcher if that plane has an error the launcher can detect or should maybe the launcher load and report an error somewhere visible in the launchers gui? — Ray St. Marie (Sep 3rd, 2015). Re: [Flightgear-devel] fgaddon c310 typos??. (powered by Instant-Cquotes)

The problem with GUI coding is that it is very, very hard - it is much more problematic than normal coding and causes far more headaches (from all the banging of your head against the wall trying to work out why it is racing and segfaulting on one system and not another). — Edward d'Auvergne (Sep 3rd, 2015). Re: [Flightgear-devel] fgaddon c310 typos??. (powered by Instant-Cquotes)

Note that as the QtLauncher transitions into a full GUI to replace the PUI GUI, many of these issues will disappear, and new ones will appear. In any case, debugging a modern GUI is much harder than debugging the non-GUI parts. A basic knowledge of threading, locks, and racing is quite useful for understanding and catching these bugs. — Edward d'Auvergne (Sep 4th, 2015). Re: [Flightgear-devel] Debug build / crash on exit. (powered by Instant-Cquotes)

One concern I have though, from the discussions over the last year, is that the PUI->Qt migration plan is not a direct one-to-one migration. I.e. to have a main Qt window providing the main FlightGear window by containing a top menu bar followed by a QtWindow GL widget underneath, in which the FG main execution loop is run as a thread from within the Qt main loop. My impression was that the GUI will be a secondary window or dialog that you have to switch to. But one problem I see with this design is that all modern GUI toolkits (Qt, GTK, metro, Cacoa, Carbon, etc.) suffer from racing and segfaulting issues if the GUI main thread or main loop/event handler is not thread number 1. — Edward d'Auvergne (Oct 13th, 2015). Re: [Flightgear-devel] GUI questions (again). (powered by Instant-Cquotes)

Canvas

Note  Building a QT based GUI is not in any way meant to obstruct the further development of a Canvas based GUI. We did voice some concerns as to whether the need to re-implement a complete widget is worth the cost is terms of development effort. The beauty of open source is that there's more than one way to do things, however. Especially in a completely modular environment (via HLA) [1].

Canvas works fine in this scenario, we can embed the texture or image a Canvas produces into a QtQuick dialog just fine. (It would be possible to reimplement Canvas using Qt but I don’t see any benefit there - the non-Qt builds still need Canvas support for flight-displays and the like) — James Turner (Oct 11th, 2015). [Flightgear-devel] GUI questions (again). (powered by Instant-Cquotes)

James isn't expecting many issues integrating Canvas layers with a Qt GUI, so there is no wasted development effort there (there is the possibility to make a Canvas backend based upon QPainter, which might be a great way to use spare CPU cycles to unload the GPU, however, as Canvas rendering becomes more intensive). However, at least for the time being, the Qt5 GUI is not even available at run-time, but will be shut down after handing over control to the main FlightGear initialization routines.

Originally, one of the main motivations for adopting Canvas and using it for UI purposes was to help getting rid of PUI (our legacy GUI engine), while coming up with a GUI that would be compatible with other important use-cases (think modern MFDs showing GUI dialogs and vice versa).

The latter is an increasingly important use-case in real life avionics (on jets/airliners) - as can be seen by ARINC 661. Such use-cases aren't easily supported by a distributed approach involving for example a browser, HTML5 and ECMAScript/JavaScript.

In addition, the idea is to keep modernizing and unifying the 2D rendering back-end in FlightGear by getting rid of legacy GL code to hopefully use modern OSG code in more and more places. At least for now, it isn't even clear if/how and when Canvas-based MFDs may support HTML5/Qt5-based UI components or vice versa (e.g. a Qt5 based dialog showing Canvas-based features, such as a MapStructure layer or MFD instances like the NavDisplay). Unlike Canvas-based efforts like the Aircraft Center, the Qt5 effort is also adding a rather significant build-time dependency for those wanting to use the new built-in launcher.

Duplicating existing functionality

While Qt5 would still be an awesome feature from a Canvas perspective, it is simply not an option as long as it is entirely optional, because at that point, some Canvas-based MFD features would no longer work for FlightGear versions without Qt5 support included. — Hooray (Jan 10th, 2016). Re: Aircraft reselection . (powered by Instant-Cquotes)

There's also the issue of code duplication and maintenance, too. As can be seen by functionality that is now getting added/re-implemented in Qt5/C++ space, despite already existing elsewhere:

• 

  This is the location tab of the new Qt5-based launcher, which is entirely hard-coded in C++

• 

  A Canvas/MapStructure based view of airports with runways/taxiways

As can be seen, there's currently no code reuse taking place when it comes to the Qt5-based location tab, and the Canvas/MapStructure-based airport/taxiway layers are very much superior in comparison, too (as well as being much better maintainable (living in fgdata)) - so it would make sense to work out a way to reuse existing code instead.

The main motivation why some people want to have a Nasal/Canvas-based GUI is to ensure that aircraft avionics can make use of the GUI toolkit, but also so that the integrated GUI can deal with existing Canvas based features (think avionics). In layman's terms, all this comes down to being able to render a MFD (PFD/ND) in a dialog box, but also to render GUI widgets to a Canvas, i.e. true recursion - as long as this use-case is supported, it does not matter whether Qt5 is used for this or not, the only thing that matters is that it would be possible to create an integrated GUI for editing/maintaining MFD avionics, but also other Canvas-feature (think having a built-in HUD or 2D panels editor). — Hooray (Jan 10th, 2016). Re: Aircraft reselection . (powered by Instant-Cquotes)

Once the PropertyBasedElement/CanvasElement wrappers are fully exposed to scripting space, we could easily register MapStructure as a new Canvas element for directly making the corresponding layers available to C++ code, without introducing any direct Nasal dependencies - i.e. the corresponding airports/runway and taxiways diagrams would be transparently created by Nasal-based Canvas elements, with the Qt5/C++ code only ever having to set a handful of properties (airport ICAO id, styling, range etc).

The main obstacle preventing us from doing this right now is that we're not yet Initializing Nasal early, so that the corresponding Nasal/Canvas bindings (via Nasal/CppBind) are not yet available when the Qt5 UI is running.

PLIB/PUI (legacy GUI)

See Howto:Processing legacy PUI dialogs using Canvas for the main article about this subject.

Up to now - only a simple replacement for the Mac Launcher has been introduced (and enabled by default only on OSX). This is not related to the internal user interface of FlightGear (which is mainly based on PUI and increasingly Canvas these days), at least for now.

However, the goal is to make an optional Qt dependency to replace PLIB PUI.

Except of course PUI will have to remain for some awkward use cases in the short term (aircraft dialogs, potentially). When compiled without Qt support, you’ll get a fully functional FG (i.e., Qt will never be a requirement to build FG), but without the GUI functions – which is indeed what some people want anyway (e.g., in some home cockpit setups).

Any in-sim GUI using Qt will exist in parallel to the PLIB/PUI one for at least one major release, although this may mean maintaining two sets of dialogs, unless the XML-translation-tool is good enough that we can use it at run-time for aircraft dialogs.

Expect to see a test branch in Git in the next month or so, and hopefully something usable for 3.6. The PLIB code will remain so hopefully it can be a gradual transition.

From what I've seen, I don't see any reason why there cannot be a direct one-to-one migration of the PUI GUI main window to a Qt main window, including dynamically generating the aircraft menu and submenus by parsing the XML unmodified. This can be done incrementally as it is developed, and probably in a way that requiresero changes to the aircraft PUI GUI XML (though changes could be made later, as you can do far, far more with Qt than with PUI). So I personally think that targeting dialogs for the PUI GUI would be the best way to go as this can be directly translated by the GUI developers into Qt dialogs requiring the aircraft developers to do nothing — Edward d'Auvergne (Oct 13th, 2015). Re: [Flightgear-devel] GUI questions (again). (powered by Instant-Cquotes)

At least for the time being, PUI doesn't seem to become obsolete anytime soon, according to James Turner's continuing preference to hack/extend and maaintain custom PUI widgets: http://sourceforge.net/p/flightgear/mailman/message/34747786/ This observation is also in line with numerous PUI related commits that he made despite his ongoing Qt5 work: http://wiki.flightgear.org/Canvas_GUI#PUI_Widgets For a more recent example, look at the waypointlist commits that are using neither Qt5, Phi or Canvas at all: http://sourceforge.net/p/flightgear/flightgear/ci/846fd2 — Hooray (Jan 9th, 2016). Re: Can't install FGRun package (+ rant about Qt5 launcher). (powered by Instant-Cquotes)

I am strongly against using a very non-standard UI (a text box + slider) to achieve that interaction. I’d much rather we add a spin-box to PUI, which is not especially hard, and give it the ‘click and drag’ mouse interaction which some spinboxes in other apps have, to give similar UI to the sliders without the visual clutter. — James Turner (Jan 7th, 2016). Re: [Flightgear-devel] Speed-up vs time-warp. (powered by Instant-Cquotes)

FGRun

One of the goals of the Qt launcher is to deprecate older frontends, like FGRun. At the time of writing (May 2016), however, it is still kept provisionally, and the user can choose whether to use the old or the new launcher. This is done for a variety of reasons:

• FGRun has more functionality than the Qt launcher;
• it is still being debated whether Qt will also be used for the internal FlightGear UI, and to what extent (especially because there is partial overlapping with the Canvas framework);
• UI integration will take some time, due to the need for legacy code to be refactored;
• some building scripts still need to be updated.wkitty42 (Sun Mar 15). Re: FGRUN I think it's useful add it in a standard way.</ref>

Python in FlightGear

See also Python in FlightGear.

See also FGPythonSys.

depending on how much progress bugman is going to make with his "python-integration-in-FlightGear" experiments (FGPythonSys), we may at some point even be able to ship FFGo as part of FGData (e.g. $FG_ROOT/Python/Apps/FFGo) and run it via FGPythonSys directly — Hooray (Feb 12th, 2016). Re: Announcing . (powered by Instant-Cquotes)

if this FGPythonSys were integrated into FG, doing things in Python that are directly committed to the core (properly reviewed---well, as the C++ code) as opposed to coming from various uncontrolled external places would certainly be pretty interesting — rominet (Feb 13th, 2016). Re: Announcing . (powered by Instant-Cquotes)

The built-in launcher that James is writing would seem to me to be a good candidate for doing things in Python (assuming there are good bindings to the C++ data structures): you need high-level access to FG aircraft and airport data, it is not performance critical, the need for threading is certainly rather limited, the code does not come from a random hangar and can be audited just like the C++ code when committed, probably by more people actually. Add to this that PyQt works very well (I have had a very pleasant experience with PyQt 3&4 under Python 2&3, and if I get enough time, I'll maybe port FFGo to PyQt---Tk is useful, clearly, but frustrating; for example Qt's QAbstractItemModel and QAbstractItemView were great to manage non-trivial tablular data; with Tk, table/tree cells are just strings and your program has to manage on its own to link that to the nice high-level Python data structures). (and I have tried other toolkits before Qt and Tk, namely GTK+ 1.2 and wxWidgets [around 2003] Bottom line: for me, Qt is the best) — rominet (Feb 13th, 2016). Re: Announcing . (powered by Instant-Cquotes)

Cross compiling using MingW64

See Building FlightGear - Cross Compiling for the main article about this subject.

As of May 2014, the dependency on Qt 5.1 makes it impossible to cross-compile the launcher using Mingw64, as the MXE project does not have makefiles for that version of QT.

Community Feedback

My issues are: I really would like to be able to see ONLY the planes I have installed on my system and not every single plane available for FG. Maybe a checkbox "show only local planes" could solve this? * I can't use a different FGDATA path, which was a very handy feature of FGRun. Even adding --fg-root=/path/to/fgdata to the "Additional options" field doesn't work. - when trying to add Aircraft and Scenery paths, I am presented with a dialog to find them, but I can't see hidden folders there, and Ctrl+H doesn't affect it I can't seem to be able to move a scenery path up and down to change the priority order, which is fundamental when using custom scenery seems to have much fewer options than FGRun, some of which were handy (for instance, where can I set texture compression now?) I love the airport view but I would expect I could position and rotate my aircraft anywhere by clicking around instead of being able to only select runways (which would be a handy feature to start at a parking position if it's not defined for a given airport) — Gilberto Agostinho (Jan 9th, 2016). [Flightgear-devel] Issues with Qt5 launcher. (powered by Instant-Cquotes)

Development

Note the launcher on next will be in flux shortly as James will add some new features - support for the aircraft package manager especially, and for specifying some directory locations to support "code only" nightly builds (see FlightGear Build Server) which use an existing FGData download.

This [add-on] screen is work-in-progress, I have moved it already in my local version, but didn’t push the changes yet. There’s new UI around the download location too, and the aircraft-dir changes you mention. For aircraft package support in general, I am hoping that setting FG_AIRCRAFT / —fg-aircraft might not be needed, i.e if aircraft-dir is explicitly set, that would be sufficient. This makes having multiple variants of aircraft co-installed much easier for the package system, but will likely need some code fixes as you already identified. — James Turner (2015-03-31). Re: [Flightgear-devel] Launcher issues on Linux. (powered by Instant-Cquotes)

this is an area I’m actively working on, especially how FG to aircraft version handling in the package system works. My current focus is to generate some test hangars on web-hosts to test against, once that works I will call out for other volunteers (Syd Adams already gracefully offered). Expect some churn in this area, in the next few weeks. — James Turner (2015-03-31). Re: [Flightgear-devel] Non-existent default catalog. (powered by Instant-Cquotes)

The problem is the initialisation sequence. The QtLauncher starts before the 'dynamic airports'. Therefore the park positions are not available (unless they have been added to the NavDataCache by previous runs of FG, and the NavDataCache is not rebuilt). — Edward d'Auvergne (Oct 22nd, 2015). Re: [Flightgear-devel] FGFS Launcher -->selecting Parkpositions ???. (powered by Instant-Cquotes)

Issues and limitations

Note that even in the optimistic case, there will be some limitations – e.g., multi-window setup will not work in Qt builds without further work.

• Airplane is in FG root, can't see in flightgear wizard
• bug #1772

The default aircraft catalog download path is http://fgfs.goneabitbursar.com/pkg/<version>/default-catalog.xml, which does not exist for 3.4.0 or 3.5.0, giving a "catalog download failure:" error and an empty Aircraft Center [ticket #1737]. — Rebecca N. Palmer (2015-03-31). [Flightgear-devel] Non-existent default catalog. (powered by Instant-Cquotes)

The "Add-ons" tab of the launcher has several problems on Linux [see posting for details] — Rebecca N. Palmer (2015-03-31). [Flightgear-devel] Launcher issues on Linux. (powered by Instant-Cquotes)

Further investigation found that the launcher *is* trying to add this directory to fg-aircraft (src/GUI/QtLauncher.cxx:772), but that this doesn't work because this option is processed before the launcher is run (intentionally, to allow the launcher to find aircraft in fg-aircraft: src/Main/main.cxx:448). — Rebecca N. Palmer (2015-04-01). Re: [Flightgear-devel] Launcher issues on Linux. (powered by Instant-Cquotes)

References