RE: de facto language forks (was: two LuaJit influences)
- From: William Adams <william_a_adams@xxxxxxx>
- To: "luajit@xxxxxxxxxxxxx" <luajit@xxxxxxxxxxxxx>
- Date: Tue, 14 May 2013 22:00:17 +0000
I've spent a fair amount of time with node/V8 and LuaJit of late. My
productivity is night and day, and it's because of the ffi. No other
language/environment/ecosystem feature has been strong enough to make this a
bad choice. Languages evolve. With PyPy using FFI, and JavaScript sure to get
the same, it seems clear to me that this is the way to go for modern languages.
It's something for the Lua maintainers to consider. I don't feel any creepy
crawlies myself. At the end of the day, I'm an engineer with no loyalties to
any language. I've used plenty over the years, and I'm sure I'll use more in
the future. I like what I see in Go, and even this new terra thing looks
interesting. I like the discussions around LuaJIT and a possible LuaJIT OS.
This is vibrancy and beneficial to me as a software engineer. I'm not so much
interested in language forks and the like because at the end of the day, I'll
use what works for me, and others can use what works for them. So, nothing
scary, there's no assured outcome that needs to be preserved. We all get to
learn and explore new things rapidly, and that's good for all of us, and for
our industry. So, let's not worry about schism and strife. We can all just code
along. The features will work themselves in a nice Darwinian way. -- William
===============================
- Shaping clay is easier than digging it out of the ground.
> Date: Tue, 14 May 2013 16:42:02 -0500
> From: drake@xxxxxxxxxxxxxx
> To: luajit@xxxxxxxxxxxxx
> Subject: Re: de facto language forks (was: two LuaJit influences)
>
> Quoth William Adams <william_a_adams@xxxxxxx>, on 2013-05-14 19:31:37 +0000:
> > As far as compatibility is concerned, I think you've already let
> > that horse out of the stable with the introduction of ffi.
>
> I'll say.
>
> I actually have nightmares about this sometimes.
>
> I have the creeping crawling feeling that PUC Lua 5.2 (or at least its
> C API and nice _ENV characteristics) is going to be dead to almost
> everyone because at some point they are going to say "well, I want to
> use $foo, which the author rewrote using the FFI, so now I need
> LuaJIT, so now I can't use 5.2, so..." or rather because anyone who
> _doesn't_ say that will find that they have to do so much more work
> that their software never gets out the door.
>
> The big dividing wedge is that the FFI loses its powers when it comes
> into contact with "classic" Lua/C functions, so it _actively_ drives
> away traditional-Lua compatibility, and even if that were magically
> vaporized it wouldn't cancel out the social motion now.
>
> I have the additional creeping crawling feeling it's going to be like
> Racket/Scheme except (or perhaps including---I haven't tracked that
> division well enough to know) that all the life may get sucked out of
> one of them, but that might be an exaggeration.
>
> Such is the human world of technology.
>
> ---> Drake Wilson
>
Other related posts:
