Mission building, this long, hard, but splendid job...

[Alfaunostebas11]

The flight simulator, for many years (I started with FS2, Commodore 64 ... how old I am!) is for me a very great passion and I have now spent tens of thousands of hours in front of the computer, with keyboards, joystick, throttle and pedal , flying in all the skies of the world, first with civil and then with military planes.
Now more than ever, having to stay home for this damned coronavirus scourge, I am full time engaged with IL-2 Sturmovik, for which (already since several mounths...) I am carrying out a campaign (on BoS), of which I have already written and tested twenty missions (the total should be around 25).
Unfortunately, with the beautiful and powerful, but very complex, editor of the new IL-2 GB, building missions is very demanding and very long.
Fortunately, the community is very helpful through the forum.
Above all, most of the time is spent testing missions, debugging, also in consideration of the fact that, unlike the old lovable IL-2 Sturmovik 46, it is not possible to quickly switch from the editor to the game to try out the changes, but every time you have to exit, launch the program, load the mission and try it.
Of course, every time you fly, you find that there are several things that don't respond as you expected, so you have to write on paper, what are the problems you have detected.
Then exit the program, go back to launch the BoS Mission Editor, study the possible changes and then try the mission again. If the changes are OK, go ahead implementing the mission with other planes, objects, objectives and so on, until you reach a result that you think can be appreciated by those who will use this work of yours ...
Personally, to get a mission that is not too simple, I need at least 20-30 versions of progressive updating (I save the version with a progessive numer after the name, so if something is wrong, I restart from the last version OK) and several working days.
So, I would like to share the best way to test with those of you who develop missions and campaigns, given that I do not know that there is a standard debugging procedure that allows you to track the progress of the program step by step while it is running , in order to understand what is the logical step to correct when something you had planned does not happen.

Another help, of course, can come from cameras placed in significant positions, or from observing the behavior of other planes / vehicles / artillery / ships etc. padlocking them during the mission. On in some occasions, it may be useful to fly the mission with navigation aids and with the icons visible on the map.

Trivial example of troubles: a Proximity trigger did not work because it had not been linked by an MCUS, or a logical action did not take place because timing was wrong, or an object had to be enabled or not, not to mention very recurring situations when bombers do not want to attack... or more complex things.
All this wastes a lot of time !!!
Aside from the intensive use of templates or groups, something that is very useful, I find that a important help is given by positioning a series of "subtitles", which warn me that the program has "passed through a certain place", or viceversa has not passed and then I go to look for the problem.
Of course it would be nice to have more help from the Editor, or perhaps a kind of log, where all the logical steps while the mission is running are recorded, or there is already a utility or something like that, but I don't know it.
So, how do you do your debugging?
Thanks, Stebas

Edited March 15, 2020 by Alfaunostebas11

[Jaegermeister]

You have just described my spare time for the past 6 months in very accurate detail.

 

That is a very thorough description of how building campaign missions works, except you forgot the breaks to work on historical skins for the aircraft, reading through detailed histories to research relevant tactics and adversaries, and writing a  cohesive storyline for the briefings.

 

Is it a waste of time, or is there just a very small group of people who get some gratification out of that process, and being able to participate in a small piece of history they have revived when things were just a little bit more “analog”?

[Alfaunostebas11]

9 hours ago, Jaegermeister said:

except you forgot the breaks to work on historical skins for the aircraft, reading through detailed histories to research relevant tactics and adversaries, and writing a  cohesive storyline for the briefings.

Bravo Jaegermeister !
I limited myself to the technical aspect of the construction of the missions, but rightly a large part of the time (for those who, like us, love historical missions) is used, preliminarily and during the drafting of each mission in historical research, of facts, documents, battle orders, departments, maps, front line, battle dates, historical episodes etc.
Each of my seven campaigns for IL-2 Sturmovik 1946 is accompanied by a manual of about 200 pages, both in Italian and in English (which is not my native language ...)!

With IL-2 GB, now, we also have the possibility to give each plane a name, so in my missions I try to use the names of real pilots who fought on that scenario ... fantastic!

Edited March 16, 2020 by Alfaunostebas11

[IckyATLAS]

As Jaegermeister said your description is basically what it is. If you just create one mission, from scratch there is not a much better or faster way to do.  If you create a scripted Campaign then the initial map configuration becomes very important as well as to very carefully define your groups. This is where it can make a big difference, when you create missions after missions in a scripted campaign. Why? simply because in a scripted campaign if you want to have a continuity from one mission to the other, with events in one mission to be continued in the next one. Like if in mission 2 you have destroyed a bridge or buildings etc. that in mission 3 they stay destroyed. Or if you have trains for troops transportation  between places, and they have to run across some missions etc. Another example is an airfield where things change over missions but you must have a continuity and evolution over time.

 

In such a campaign type if you are not very careful in having well designed things, then the time lost can become huge. This means that if you do a change in mission 5 that has to be reflected back in Mission 1,2,3 and 4, then this means you have to modify all those missions, and the modifications can be very time consuming and have some unwanted side effects, which means retesting those missions. And if you have not designed things properly, then it becomes an enormous amount of work. 

 

 

 

[Flashy]

not too much to add to what has already been said, but you can actually run the mission editor and the game at the same time, although it will often create weird visual glitches if you go into "first person" mode in the editor (f9) when the game is running. As long as you avoid pressing f9 in the editor though, I find you can run the two side by side without too much trouble, and this makes editing and quickly checking the result faster. Even if the game does glitch out, you can still test certain things because it doesnt affect things like subtitles or mission logic, only the visuals..

[Alonzo]

I use subtitle MCUs named DEBUG which print text in bright green text at the bottom of the screen. This tells me what is happening, for key bits of logic (triggers etc). I then use a 'strip debug' Perl script to set the coalition for the DEBUG messages so no-one can see them when I run the mission on a multiplayer server.

 

Another trick is that I save my work into zip files. I just zip everything up and name it 0.zip, then 1.zip, and so on. That way if I truly screw something up, I can go back to a zip file. I make a new zip every half hour or when I've done something significant. I Ctrl-S in the editor compulsively, as I find it crashes quite often.