From: Chad Perrin <code@...>
Date: 2011-07-21T23:19:29+09:00
Subject: Re: Ruby in JavaScript, all projects are dead?

--WYTEVAkct0FjGQmd
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Thu, Jul 21, 2011 at 01:57:31PM +0900, David Masover wrote:
> On Tuesday, July 19, 2011 05:41:47 PM Alexey Petrushin wrote:
> > UPD, this one seems to be interesting and is very active. And (a big
> > plus), it generates human readable js.
> >=20
> > https://github.com/adambeynon/opal
>=20
> I remember starting a project like this. I more or less gave up when I=20
> realized there didn't seem to be a good way to have it continue to genera=
te=20
> human-readable JS, and also support continuations.
>=20
> This seems to suffer from the same limitation. Notice there's no enumerat=
ors.=20
> In MRI, I can do stuff like this:
>=20
> ruby-1.9.2-p180 :010 > (1..10).each_cons(2).to_a
>  =3D> [[1, 2], [2, 3], [3, 4], [4, 5], [5, 6], [6, 7], [7, 8], [8, 9], [9=
, 10]]=20
>=20
> Opal doesn't do that -- all of the enumerable methods (each and friends)=
=20
> simply die if I don't pass a block.
>=20
> I guess if that ends up being the only edge case, and people are OK with =
this,=20
> it makes sense. But it seems like the more weird corner cases like this, =
the=20
> less benefit there is to trying to pretend it's a full-fledged Ruby. Even=
 if=20
> your gems don't hit any of the corner cases imposed by the interpreter it=
self,=20
> you're still running in a restricted environment that may pose its own=20
> problems...

It would be nice to have a full-fledged Ruby that compiles to
human-readable JavaScript, but if it falls short of that level of
completeness, that does not necessarily make it unworthy of developing
and supporting (and, most importantly, of using).


>=20
> Maybe it would be better to create a language specifically designed to wo=
rk=20
> well in that environment -- to build on the good ideas from JavaScript, g=
ive=20
> us Ruby-esque syntactic sugar, but to stick close enough that the resulti=
ng=20
> JavaScript makes sense and is debuggable... and I just described CoffeeSc=
ript.
>=20
> So, it looks really, really cool, but I think I'm still going to use=20
> CoffeeScript if I use anything beyond pure JS.

That might make sense.

What I'd *really* like, though, is a fixed JavaScript.

--=20
Chad Perrin [ original content licensed OWL: http://owl.apotheon.org ]

--WYTEVAkct0FjGQmd
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.14 (FreeBSD)

iEYEARECAAYFAk4oMYUACgkQ9mn/Pj01uKXwxQCfVe1aWY+lPrkprUrAR9Vc6Xu3
TSYAnjoTW0kKLPQ/JnvC+XD60knNWX8n
=zBkd
-----END PGP SIGNATURE-----

--WYTEVAkct0FjGQmd--