DutchPilotMitchell wrote in Wed Sep 07, 2011 5:49 pm:Great, A-10 scenario's work, but the ferry scenario's don't work. No offense but, do you actually test the scenario's before releasing them?
Thanks for the bug report, the paste of the errors was especially helpful.
Bombable is a fairly large and complex package now. Doing real regression testing of each point release across all different aircraft, AI aircraft, scenarios, platforms, and at least a few different version of FG as well as the current GIT version, and each of those on multiplayer vs AI, is basically impossible without a fairly good-sized team to work on it. (Keep in mind that you have to multiply all those options together to get your total number of test cases--ie, you might have an error that appears only for one certain aircraft in one certain scenario, on a certain version of linux, and version 2.2.0 of FG. And oh yeah--only if FG's MP is enabled but Bombable's MP is disabled.)
In short, no I don't methodically check each and every aircraft, each and every scenario, each and every AI aircraft, and each and every aircraft over MP, before putting out every point release. If I tried to do that, point releases would be years apart . . .
In addition, once the final release/install file is created, one really should test each aircraft/AI aircraft/scenario/platform etc. by installing it on a clean install of FG. So now we're talking double that many years apart . . . .
So the reality is if we want to have releases at all they are going to be less than perfect in all the ways one could predict--like this one, which I believe was caused by some malfunction in zipping up the final files and some directory naming problems that don't show up on my platform (Windows, famous for not being able to discriminate B from b . . . ) no matter how much I test.
Thanks to your report, I think (?!) I have it fixed now. So, I really appreciate i!