More bug fixes

When Comments go bad

More progress to report this week (will update weekly till we release). The broken input functions are no longer broken and the Apache is now airworthy. I'm to blame for not paying attention to how I use comments in code. (I'd hard coded the device index during a debug session a long long time ago and commented out working code...for some reason.) I know at least one person who will be disappointed by my use of comments. There's a school of thought that suggests comments serve no practical purpose in source code. Code should be written to be fully understandable without comments in the first place therefore comments have no place in a source file. Not sure I fully agree with this but since I got stung by commented code it's one I'll look at more carefully in future. Another lesson learned.

Next up to fix...stores

Now that controllers are working again it time to go back to the stores to finish implementation of the rockets and Hellfires. My effects for building destruction are simple at best, could certainly use more pyros, especially on the gas station.

The soft-particle shaders are darker than they should be. That's one to investigate (possible suspect is missing terrain layer which might be used to modify the particle color). In which case it's just an art problem.

Macklebee (aka Mike) from the Leadwerks community has graciously donated some of his time to go over the Apache LUA script. He found a number of issues as well as some oddities in the Leadwerks engine. Since Leadwerks 2.51 isn't being supported anymore we'll have to try and find ways to work around them. These include joints that spawn at the world origin and creep towards their set location. Another problem seems to be having aircraft joints in contact with the ground when spawning them. Any collisions that occur before joints are set creates numerous problems. Seems the trick is to set the vehicle mass to zero during initialization, wait for a *single* physics update and THEN set the vehicle mass. This should fix the problem. 

Mike is welcome to code review everything from now on.

Work Items this week...to fix:

• Rocket ballistics
• Hellfires and MFD pages to finish
• Bobup-mode HUD
• Crew event speech
• Assorted polishing issues (particles, range events, rain/weather)

Facebook

I'm now a regular visitor to our Facebook presence if you want to engage in idle banter or ask questions that don't involve release dates (as according to my wife I exist in a separate space-time continuum) then please stop by https://www.facebook.com/CombatHelo

Parting thanks

Finally I want to thank again Macklebee and Klepto2, both from the Leadwerks community. Mack for his help with some bugs and Klepto2 for assistance with his extended NET DLL which exposes and adds much needed functionality in the Leadwerks engine. 

More updates soon.

Update on the Audio

Adding crew speech is a lot of fun. It adds so much to the feel of flying. At the same time it's quite labour intensive and I've discovered a lot of short-comings that I hadn't anticipated.

For example; pilot voice types. You have one set of recordings so you only have one pilot. But what if I want more, or modders want to add different voices for front seat and back seat? Then I need to pass not only the phrase but also a pilot ID or index. And what if the phrase is something that is repeated....a lot? And it's repetitive to the point of annoying? So a single phrase may have a choice of multiple speech samples. All of this needs to go into a small data file. But I'm not doing that now, instead I'm keeping this simple and allow room to 'big it up' later.

I added a frame-hook when the TGameAudio class is instantiated. The hook listens for any crew speech that's been cued or playing. When it detects that speech is ready or completed it also plays background static and a roger bleep effect. I'm pretty sure internal comes doesn't have a tx bleep but you often hear background noise. I added a tx bleep to test it and it sounds better than I expected. I now need to find someone with a Californian or a Texan who can supply me with a never-ending flood of phrases, there's always room for more or even variations. At an average of 7k of memory per phrase there's room for more.

Final audio is saved as an OGG file at 8,000 samples.