Concepts

  1. Git and GitHub
  2. Branches: master & develop
  3. Build result vs. source files
  4. Static loading
  5. Dynamic loading of mission scripts
  6. Dynamic loading of Moose
  7. Automatic dynamic loading
  8. IDE vs. Notepad++

If you want to get deeper into Moose, you will encounter a few terms and concepts that we will explain here. You will need them for the later pages.

Git and GitHub

Moose has about 260.000 lines of code and the amount is increasing each week. To maintain such a big code base a vcs (version control system) is needed. Moose uses GitHub as developer platform to create, store, and manage the code. GitHub uses Git as version control system and provides additional functionality like access control, bug tracking, feature requests and much more.

As a Moose user you don’t need to learn how to use Git. You can download the files on GitHub with a browser. But using Git will ease up the steps to keep the Moose version on your hard disk up to date.

You will need to interact with GitHub. At least to download the Moose files. For non-developers the page can be confusing. Take your time and read this documentation. We are not able to explain every single detail on using GitHub and Git. Especially because it is changing really quick and this documentation will not. So try to use the help system of GitHub or find some videos on YouTube. If you get stuck ask for help in the Moose Discord.

Moose uses more than one repository on GitHub which doesn’t exactly make it any clearer. A list can be found on the reposities page.

Branches: master & develop

As already explained in the overview two branches are used:

  • Branch master: Stable release branch.
  • Branch develop: Newest development with more OPS classes.

As a starter it is okay to begin your journey with the master branch. If you are interested in some newer classes you need to use the develop branch. The later one is also very stable, but it’s missing more detailed documentation and example missions for some of the new OPS classes.

You can switch between these branches with a drop down in the upper left corner of the GitHub repository page. The list of branches is long. So it is a best practice to save a bookmark in your browser with the links above. Both branches are available on most of the different repositories. But because of a limitation of GitHub pages, we had to split the documentation in two different repositories:

Build result vs. source files

Moose consists of more than 140 individual files with the file extension .lua. They are places in a directory tree, which makes it more organized and its semantic is pre-defined for IntelliSense to work.

On every change which is pushed to GitHub a build job will combine all of these files to a single file called Moose.lua. In a second step all comments will be removed to decrease the file size and the result will be saved as Moose_.lua. These both files are created for users of Moose to include in your missions.

The individual .lua files are used by the Moose developers and power users. It is complicated to use them, but in combination with an IDE and a debugger it is very useful to analyze even complex problems or write new additions to the Moose framework.

Static loading

If you add a script file with a DO SCRIPT FILE trigger, like we described in Create your own Hello world, the script file will be copied into the mission file. This mission file (file extension .MIZ) is only a compressed ZIP archive with another file ending.

If you change the script file after adding it to the mission, the changes are not available on mission start. You have to re-add the script after each change. This can be very annoying and often leads to forgetting to add the change again. Then you wonder why the mission does not deliver the desired result.

But when the mission is finished you can upload it to your dedicated DCS server or give it to a friend and it should run without problems. This way of embedding the scripts do we call static loading and the resulting mission is very portable.

Dynamic loading of mission scripts

The other way of loading scripts is by using DO SCRIPT. This time the mission editor don’t show a file browse button. Instead you see a (very small) text field to enter the code directly into it. It is only useful for very small script snippets. But we can use it to load a file from your hard drive like this:

1
2
3

dofile('C:/MyScripts/hello-world.lua')
dofile('C:\\MyScripts\\hello-world.lua')
dofile([[C:\MyScripts\hello-world.lua]])

So all lines above do the same. In Lua you need to specify the path with slashes, escape backslashes or use double square brackets around the string. Double square brackets are usefull, because you can copy paste the path without any modification.

If you upload a mission with this code, you need to create the folder C:\MyScripts\ on the server file system and upload the newest version of hello-world.lua, too. The same applies, if you give the mission to a friend. This makes the mission less portable, but on the other hand the mission uses the file on the hard disk, without the need to add it to the mission again. All you need to do is save the file and restart the mission.

The following can be used to increase portability:

1

dofile(lfs.writedir() .. '/Missions/hello-world.lua')

The function lfs.writedir() will return your Saved Games folder. So you place the scripts in the subfolder Missions. This way the folder structure is already available on all target systems. But you need to ensure mission and script are both in sync to avoid problems. If you changed both and upload only one of them to your server, you may get trouble.

There is another method you may find useful to dynamically load scripts:

1
2
3

assert(loadfile('C:/MyScripts/hello-world.lua'))()
assert(loadfile('C:\\MyScripts\\hello-world.lua'))()
assert(loadfile([[C:\MyScripts\hello-world.lua]]))()

It is a little bit harder to read and write because of all these different brackets. Especially the one on line 3. But it is a little safer than dofile. Because of readability I prefer to use dofile.

Dynamic loading of Moose

Of course you can use the same method to load Moose. This way you can place one Moose file in your Saved Games folder, which is used by multiple missions. If you want to update Moose you just need to replace the file and all missions will use the new version. But I prefer to add Moose by a DO SCRIPT FILE trigger so I can add and test the new version for each mission step by step.

But we added two different ways to load the Moose source files automatically. This is useful for Moose developers and it is a requirement to use a debugger. This will be explained later in the Debugger Guide.

Automatic dynamic loading

With the code below you can have the advantages of both approaches.

  • Copy the code into your mission script at the beginning.
  • Save the mission script into the folder Missions in your Saved Games folder.
  • Change script filename in line 2 to match to your script.
  • De-Sanitize your MissionScripting.lua.

Now the mission will use the script on your hard drive instead of the script embedded in th MIZ file, as long as it is available. So you can chnge the script, save it and restart the mission, without the need to readd it after each change.

If you reach a stable state in your script development and want to upload the mission to your server or give it to a friend, then just add the script again like in the static method and save the mission.

Do not forget to readd the script, prior uploading or sharing the mission, or it will run with an outdated version of your script and may fail if the objects in the mission don’t match to this old version.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22

-- Use script file from hard disk instead of the one included in the .miz file
if lfs and io then
  MissionScript = lfs.writedir() .. '/Missions/hello-world-autodyn.lua'
  -- Check if the running skript is from temp directory to avoid an endless loop
  if string.find( debug.getinfo(1).source, lfs.tempdir() ) then
    local f=io.open(MissionScript,"r")
    if f~=nil then
      io.close(f)

      env.info( '*** LOAD MISSION SCRIPT FROM HARD DISK *** ' )
      dofile(MissionScript)
      do return end
    end
  end
else
  env.error( '*** LOAD MISSION SCRIPT FROM HARD DISK FAILED (Desanitize lfs and io)*** ' )
end

--
-- Simple example mission to show the very basics of MOOSE
--
MESSAGE:New( "Hello World! This messages is printed by MOOSE!", 35, "INFO" ):ToAll():ToLog()

IDE vs. Notepad++

As a beginner you should start with a good text editor, which supports syntax highlighting of Lua code. This must not be Notepad++. It can be any other powerful editor of your choice. Do yourself a favor and don’t use the Windows editor.

If you are a developer of Lua or another programming language, then your are most likely familiar with an IDE (Integrated Develop Environment).

Otherwise you should know, that an IDE may help you with code completion, Refactoring, Autocorrection, Formatting Source Code, showing documentation as popup on mouse hover over keywords and Debugging.

There are different IDEs available. And not all IDEs support all features. The three most important for Moose are:

Eclipse has the best support for hover documentation and IntelliSense with Moose. The Inventor of Moose (FlightControl) did an amazing job by adding an integration to Eclipse LDT (a special version for Lua). Unfortunately Eclipse LDT is not maintained any longer (last release 2018). And the debugger doesn’t work anymore, since an update of DCS.

In Visual Studio Code the support of Lua can be added by an addon. The debugger works with Moose and DCS, but showing the LuaDoc and IntelliSense is very limited.

PyCharm supports Lua also with an addon. The debugger works with Moose and DCS, but showing the LuaDoc and IntelliSense is very limited.

It is up to you to choose the IDE according to your taste. Guides on how to setup Moose with different IDEs and Debuggers are provided later in this documentation.