GSOC applications coming up

GSOC applications coming up

[No Itisnt]

Side note: For those of you who don't follow the libgc list, libgc is
now patched to work with Valgrind. Checkout the CVS version, compile
with -DUSE_GET_STACKBASE_FOR_MAIN and valgrind away.

I'm sorry I wasn't able to contribute anything yet (although I did in
a roundabout way). At the moment I'm tinkering with the arithmetic
parser I posted to guile-user, and trying to turn it into a parser
generator.

I expressed some interest in doing GSOC with Guile a little earlier
and if y'all are willing to back me up, I really should write a
proposal and submit it (I guess I'd just submit it to the GNU Project,
assuming you approve?) I'm still plenty open to suggestions; I'm
particularly interested in parsing and compilation, but mostly I just
want to write some Scheme.

What exactly needs to be finished in the Javascript compiler?

Let me know what you think.

Re: GSOC applications coming up

[Andy Wingo]

Heya,

On Mon 29 Mar 2010 00:17, No Itisnt <theseaisinhere@gmail.com> writes:

> At the moment I'm tinkering with the arithmetic parser I posted to
> guile-user, and trying to turn it into a parser generator.

Cool!

> I expressed some interest in doing GSOC with Guile a little earlier
> and if y'all are willing to back me up, I really should write a
> proposal and submit it (I guess I'd just submit it to the GNU Project,
> assuming you approve?)

Sure, that would be fantastic.

> I'm still plenty open to suggestions; I'm particularly interested in
> parsing and compilation, but mostly I just want to write some Scheme.

How about that Lua implementation you proposed? I don't think there's
enough left on the Javascript compiler for it to be a summer project,
but poking around in that code could be instructive to see how it all
breaks down between the parser/compiler and the GOOPS-based runtime.

> What exactly needs to be finished in the Javascript compiler?

The standard library, mainly. I left it in a poor state, unfortunately,
without test cases, and without a TODO list; I am a bad person :P

> Let me know what you think.

I think it's going to be a lovely summer :) Speaking from experience,
writing compilers and runtimes is really, really educational, and
often the results are practically useful. It would be great to be able
to extend Emacs in Lua.

Cheers,

Andy
-- 
http://wingolog.org/

Re: GSOC applications coming up

[No Itisnt]

Hi, I don't think I can reply on the GSOC site, so I'll put it here

> Specifics on your proposal:
>
> * cothreads: Guile's partial continuations can implement this construct well, and correctly; while it's not necessary, it would be lots of fun. See "Prompts" in the manual.

I left it out because I have never used continuations. I'm confident I
will finish the entire standard library but the best laid plans of
mice and men and all that

>
> * repositories: I'd personally prefer to just have you commit to Guile's repository, in a branch, if you're happy with that.
>

It's all the same to me

> * test cases: Ideally you'd write something that uses Guile's internal (test-suite lib), adding lua.test to test-suite/tests/.
>

Yes

> * parser: how do you plan to implement this? I would use a lexer and an LALR parser. We will have an LALR parser generater documented and in core within the week.
>

Well I can tell you now I don't want to write it by hand. If my TDOP
parser bears out, I'll probably use that, otherwise I'll be looking
into parse-lalr etcetera. I'm also excited about PEG but it may not be
ready in time :)