New HaxeFlixel is late, and I have to take a decision
I was waiting for the new version of HaxeFlixel finally supporting the latest Openfl and Lime, but time is running and I have to choose the framework for my project: if I decide to take HF will I have to consider a partial rewrite of the code when the new HF will be out or can I sleep well because the passage will be painless?
New version is developed publicly, discussion here: https://github.com/HaxeFlixel/flixel/pull/2068
You should try it already, developers know nothing about your project and don't have it on their computers to test if it would be broken
Since 3.6.1 in legacy mode a lot has changed in OpenFL, and my project got some regressions in rendering
Ok, I can try it, but what I wanted to know is how much the next HF will be "different" from the current official one: will it be different under the hood only or will the way we code games change too? I think there will be (a lot of) new functions too, but will the old ones continue to work the same?
I suppose that if I start my project on the dev version I won't care about the differences with the current version, right? ;)
...as simple as nuke a building and remake it from scratch...
Mostly under the hood, but it becomes GPU-accelerated, so you shouldn't, for example, call copyPixels each frame, but use shaders instead to get performance
because of Tiles?
Because it would be the slowest part of rendering. My example may not be the best, because copyPixels in new versions of OpenFL is really slow anyway, so you don't want to use it unless you are ready to do some optimization work on it
Note that I'm talking about native targets, flash target is mostly unchanged AFAIK