Re: [open-cobol-list] Daydreaming about another language to emit

[Brian Tiffin]

Charles;

Oops, didn't mean to ruin a good one-liner.   ;-)

When I posted the LLVM clang success story on the #clang IRC channel,
there were a few folk there that said they'd like to see a version
that emitted LLVM bytecode.  Someone even ran off the chat to see how
hard that would be to pull off.  I never heard back, but I never
really poked back into that chat room.

Cheers,
Brian

On 10/5/13, Charles Anthony <address@hidden> wrote:
> On Sat, Oct 5, 2013 at 8:14 AM, Brian Tiffin <address@hidden> wrote:
>
>>
>> Charles;   LLVM.  We do clang.  Build compiler with clang, generated
>> code compiled with clang.  All worked first try.  One small compiler
>> switch change meant ALL the make check passes as well.
>> http://opencobol.add1tocobol.com/#does-opencobol-work-with-llvm
>
>
> As the original poster was discussing removing the C interface altogether,
> I was positing COBOL --> LLVM directly. In an abstract sense this is good
> as LLVM is targeted as a compiler back-end, as opposed to C which is
> powerful enough to be used as a compiler back-end.
>
> Be clear, I am not criticizing OpenCOBOL's implementation, I am just doing
> what-if games. If I was tasked with with the job of reworking OpenCOBOL
> with the express intent of replacing the C back-end, I would advocate LLVM.
> (To some extent because I've used it a write compilers).
>
> Also, the original question  seemed to seeking more information about the
> engineering aspects of different approaches, so I was recommending that he
> read about LLVM to better understand the compiler back-end problem space.
>