What do you want to hear in an OpenFL "Tips, Tricks and Advanced Use" session? (Haxe Summit, May 3-5 2018)

I am preparing two talks for Haxe Summit, which is going to be May 3-5 in Seattle this year.

One of the talks will cover the history, present and future of OpenFL, but the other is slated to be more of a tips, tricks or advanced use session. First, I’m wondering if this seems like a good divide of time (suggestions are welcome) but I’d also love to hear what things you would want to hear more about. I’ve heard multiple mention using SWF assets, but more specifics or other suggestions would be really helpful.

What do you want to hear?

Thank you for your help!

Not sure if that fits, but I’d looove to hear/read about openFL GPU rendering.

I’m coming from as3/Starling land, never used regular display list much, so I’m used to be careful with draw calls and textures. I read that openFL uses GPU rendering when possible, but how does it work ?
Is it smart/optimized, able to share textures between objects, to minimize draw calls ? Does it create texture atlases internally or something ? Are there guidelines/tips to follow to help with performance ? When to use regular openFL and when better use Starling ?

Also : text. One thing that bothered me with Starling and Feathers (UI lib) is that text needs to be rendered to a texture (except for bitmap text of course, but it’s not efficient for relatively long texts), which is pretty slow in flash and costy draw call - wise. How does GPU accelerated text work in openFL (if it is GPU-accelerated at all ?) ?

Another thing I’d like to hear/read about is the new “Packed Asset Library Support”, although I will probably look into it when I get some time : does it work for non-embedded assets ? If yes what’s the proper way to load them ? Assets.loadLibrary (and .unloadLibrary to clear it then ?) ? Any drawbacks ? What are the packed assets IDs like ?

1 Like

More information on how to make native extensions. There is little documentation about it.


I would like to hear about your top non-flash-api features of Lime/OpenFL and what’s your plan for the future of the framework.

1 Like

I’d love to hear about future features as well when it comes to cross platform app functionality (giving Javascript a run for its money :smile:)

Create a simple project in one of Adobe’s tools, and then show the fastest way to create a new OpenFL project based on that that compiles to a non-flash target. And then how to keep it updated.


Document what works well and doesn’t in terms of performance: manage expectations.