Hello,
On Fri, 27 Feb 2004 17:10:44 +0100, Drew McCormack
<[log in to unmask]> wrote:
>On Feb 27, 2004, at 4:15 PM, Aleksandar Donev wrote:
>
<snip>
>> Many and many years of development have gone into Fortran compilers.
>> Developing all the optimizations just for, for example, array
>> operations,
>> takes time and effort. Any competition would need a long time to catch
>> up.
>Not at all. None of that would be lost. Any new language could simply
>be built on top of the current fortran backend. Any new language should
>basically include Fortran 90's nice array capabilities; they are the
>state of the art. Fortran falls down somewhat in other areas in my
>view, and it is there that improvements could probably be made.
Most vendors today use separate front-ends for each language,
and a common code generator. The difference is in what constraints
each front-end puts on the code generation, in no small measure,
when the front-end can guarantee that memory is not aliased.
Note that claims that "all languages are subject
to equal optimization" (which I've actually heard), are, at best,
claims based upon a theoretical compiler, compiling the entire
application in one go, and having unbounded time
and memory to compile.
--
Cheers!
Dan Nagle
Purple Sage Computing Solutions, Inc.
|