the huge question which hunts me at the moment -> on what part i should work today.
here some points i want to implement soon, some are important for the release at june, some arnt:
- depthmap related collision system (only runs on the gpu)
- particlesystems (cpu based ones and gpu ones)
- webcl ( which could be used for the rudimentary physic implementation, either particles, and everything else which could be good for multiprocessing.) but therefor you have to download something. due this point, im not so happy with.
- maya model exporter
- LSD model pipeline
- input devices
and tons of other stuff...
oh and id forgot about an very necessary part, the unliked process to write an documentation for. but this point is at the moment senseless. as far LSD isnt settled and fixated in all the pipelines and how to use it, it wouldnt make any sense to start with that point.
so, it could happen -> a release with an unready documentation.
hmpf, seems there so many points which needs to be done. but im going to focus on the game and release related things, such as sound, input devices, particlesystems...
so changed the plan a bit. isnt it, that i wanted to work on the "circle of dead" game :D? after the LSD v.0.2 release in middle of june, im going to concentrate me (at my private solo projects (LSD and circle of dead)) onto this game again.
after couple of hours at the dreamland (sleep), i had to do a decision about the vbo problem i had. i fixed this problem by using a workaround which i tried to avoid before. the good thing now it works, the bad thing - im sure i will spend into this problem again alot of time. ;)
but in overall it makes sense to use now this quick and dirty solution. as long im a rookie in opengl programming, it would take ages to solve this problem by using a elegant way for. when im getting deeper into this matter, im sure this problem will be fixed in half time or less. so for now im able to concentrate me onto different parts at LSD. and as far you can imagin, lot of things needs to be done or improved.
stay tunend.. smooth your way n´ keep the eyes on the rabbit hole
actually i wanted to fix an bug in my render pipeline -> without any success. hopefully i get it tomorrow working...
so now im playing with the idea, to stay off of the 3D geo implementation till the middle of june and focusing for now on the 2D part which is basicly the main issues for this engine. i will see what the next day brings - if its the same like today, im getting for a short time a bit on distance to the 3D geo.
well, the problem is, if i want to render several pieces into on render to texture (rtt) i do access the vertexbuffer (vb) from the object before. its a bit of a pain. i do know some work arounds to handle this problem. but all this solutions feels that dirty.