using loadanimmesh for the level mesh slows down the app afaik. further one has to parse the whole family tree to check any linepick or collision. I vote for a static level mesh without labels handling. It seems to be a nice idea to be able to name a child "put a light here", but in the end it's way more "WYSIWYG" when you simply position a Light in a Editor written in Blitz.
Yes, of course, I can offer my Editor. But this would mean we will have about the same possibilities like the csp engine. My wrapper didn't handle every possible setting. Although I think it's enough. This way we'll have a level loader too. But - you know, it's written in my style. No types in use. It seems they were not required :)
BTW. I added your seashell idea (automated for the whole level), and put the A-Star onto the waypoints it produced (was dammit tricky). Now it works 90%. I still got to fix some little bugs and solve some logical problems.
There is still an unbearable lag when it searches a path to a location that is not on the same floor. Maybe this should be solved with predefined paths to the next floors.
Did you like the level? Took me a whole day.
Yes, of course, I can offer my Editor. But this would mean we will have about the same possibilities like the csp engine. My wrapper didn't handle every possible setting. Although I think it's enough. This way we'll have a level loader too. But - you know, it's written in my style. No types in use. It seems they were not required :)
BTW. I added your seashell idea (automated for the whole level), and put the A-Star onto the waypoints it produced (was dammit tricky). Now it works 90%. I still got to fix some little bugs and solve some logical problems.
There is still an unbearable lag when it searches a path to a location that is not on the same floor. Maybe this should be solved with predefined paths to the next floors.
Did you like the level? Took me a whole day.