5 ms·
The upshot is likely that a community-created Lua or JS binding will gain a significant userbase.
by leoc 13y ago
The upshot is likely that a community-created Lua or JS binding will gain a significant userbase.
- TillE 13y agoThat's fairly likely, and there are several of those projects underway in the forums. Unfortunately the engine is filled with weird macros and their own reimplementation of STL-like types, so any full binding is going to require quite a lot of wrapping. I think the more effective solution for most developers will be to keep all the engine-adjacent code in C++, but integrate a scripting engine of your choice just with your game logic.
- CyberShadow 13y agoNote that you'll need a language capable of AOT compilation if you want to target consoles which only run signed code (or use an interpreter, which will be rather slow). JIT is not an option on those systems.
- dkersten 13y agoThe LuaJIT interpreter is surprisingly fast. Still not as fast as the JIT or an AOT compiled language though...
- NickPollard 13y agoHis point though is that JITs (like LuaJIT) are often not allowed on consoles for security reasons - they don't allow running of unsigned code. So LuaJIT might not be an option.
- myrmidon 13y agoLuaJIT contains a pretty fast interpreter, thus it makes sense to use it on consoles (with the JIT turned off by compile-time-flags) instead of the Lua reference implementation.
- dkersten 13y agoAs myrmidon said, LuaJIT ships with a very fast interpreter. What I was trying to say is that, while its not as fast as JIT or AOT, it is still extremely fast (many times faster than the reference Lua implementation apparently).
- CmonDev 13y ago"JS" - they want a better language, not worse one.