Requesting Feedback: I Love Haskell, but can't find a place to use it - Haskell-cafe

The old Google Groups will be going away soon, but your browser is incompatible with the new version.
Message from discussion Requesting Feedback: I Love Haskell, but can't find a place to use it
Received-SPF: pass (google.com: best guess record for domain of haskell-cafe-boun...@haskell.org designates 78.46.100.180 as permitted sender) client-ip=78.46.100.180;
MIME-Version: 1.0
In-Reply-To: <CAFnw-00+NFMp61p9xb=1RB84sSmh_gZ3AuowZfm5gXpdScg...@mail.gmail.com>
References: <CAFnw-00+NFMp61p9xb=1RB84sSmh_gZ3AuowZfm5gXpdScg...@mail.gmail.com>
Date: Wed, 30 May 2012 22:32:08 -0400
Message-ID: <CAKFCL4UYA53gO-CSYunJrLecsrZ7aivYOWUniWRvCHrdXjn...@mail.gmail.com>
From: Brandon Allbery <allber...@gmail.com>
To: Jonathan Geddes <geddes.jonat...@gmail.com>
Cc: haskell-c...@haskell.org
Subject: Re: [Haskell-cafe] Requesting Feedback: I Love Haskell,
 but can't find a place to use it
Precedence: list
Content-Type: multipart/mixed; boundary="===============7871007626186561052=="
Sender: haskell-cafe-boun...@haskell.org
Errors-To: haskell-cafe-boun...@haskell.org

--===============7871007626186561052==
Content-Type: multipart/alternative; boundary=e89a8fb1f182d38d2d04c14bdf83

--e89a8fb1f182d38d2d04c14bdf83
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

On Wed, May 30, 2012 at 8:30 PM, Jonathan Geddes
<geddes.jonat...@gmail.com>wrote:

> I had hoped that compiling Haskell to C with -fvia-C (or would it be just
> -C?) would allow Haskell to run in new, uncharted territory such as Andro=
id
> (with NDK), IOS, Google's NaCl, etc. But today I learned that GHC's C
> backend has been deprecated!  Is it more difficult than I am imagining to
> get Haskell to work in these environments? Is it simply a matter of low
> interest in this kind of work? Or something more fundamental? Am I missin=
g
> something?
>

The C backend was never suitable for that; it couldn't cross-compile and it
used some crufty Perl to apply dubious optimizations to the generated code
(and never did anything interesting on x86 anyway due to lack of
registers).  The portable ANSI C backend, which is used to port GHC to a
new platform in the absence of cross-compilation, still works =E2=80=94 but
produces rather slow code.  The native and LLVM backends are much better,
and there is at least some potential for cross-compilation.

I believe there is work proceeding on an ARM native code generator that can
be used to target Android.

--=20
brandon s allbery                                      allber...@gmail.com
wandering unix systems administrator (available)     (412) 475-9364 vm/sms

--e89a8fb1f182d38d2d04c14bdf83
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">On Wed, May 30, 2012 at 8:30 PM, Jonathan Geddes <span dir=
=3D"ltr">&lt;<a href=3D"mailto:geddes.jonat...@gmail.com" target=3D"_blank"=
>geddes.jonat...@gmail.com</a>&gt;</span> wrote:<br><div class=3D"gmail_quo=
te"><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-lef=
t:1px #ccc solid;padding-left:1ex">
<div>I had hoped that compiling Haskell to C with -fvia-C (or would it be j=
ust -C?) would allow Haskell to run in new, uncharted territory such as And=
roid (with NDK), IOS, Google&#39;s NaCl, etc. But today I learned that GHC&=
#39;s C backend has been=C2=A0deprecated!=C2=A0=C2=A0Is it more difficult t=
han I am imagining to get Haskell to work in these environments? Is it simp=
ly a matter of low interest in this kind of work? Or something more fundame=
ntal?=C2=A0Am I missing something?</div>
</blockquote><div><br></div><div>The C backend was never suitable for that;=
 it couldn&#39;t cross-compile and it used some crufty Perl to apply dubiou=
s optimizations to the generated code (and never did anything interesting o=
n x86 anyway due to lack of registers). =C2=A0The portable ANSI C backend, =
which is used to port GHC to a new platform in the absence of cross-compila=
tion, still works =E2=80=94 but produces rather slow code. =C2=A0The native=
 and LLVM backends are much better, and there is at least some potential fo=
r cross-compilation.</div>
<div><br></div><div>I believe there is work proceeding on an ARM native cod=
e generator that can be used to target Android.</div><div><br></div></div>-=
- <br>brandon s allbery =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0<a href=3D"mailto:allber...@gmail.com" target=3D"_blank">allbery.=
b...@gmail.com</a><br>
wandering unix systems administrator (available) =C2=A0 =C2=A0 (412) 475-93=
64 vm/sms<br><br>
</div>

--e89a8fb1f182d38d2d04c14bdf83--


--===============7871007626186561052==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Haskell-Cafe mailing list
Haskell-C...@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe

--===============7871007626186561052==--
Create a group - Google Groups - Google Home - Terms of Service - Privacy Policy
Account Options
