[MLton-devel] Re: MLRISC

Lal George lg@network-speed.com
Tue, 14 Jan 2003 11:43:09 -0500

I lied ...
The RegisterAllocator functor is higher-order -- curried, but
AFAIK there are none of the other non-standard features you

I suspect that the most common use of the RegisterAllocator
functor is one where it can be fully applied at one place --
that is to say, we don't carry around the partial application.

The most recent version of MLRISC is always the one bundled with
the compiler, as they share the same CVS repository.

The library has a lot of dead code, so it is not surprising
that many of the .cm files do not compile. A lot of the dead
code were experiments that worked at one time, but abandoned
for one reason or another.

At some point Allen and I will need to do a significant clean


Matthew Fluet wrote:
> Lal,
> Stephen Weeks mentioned to the MLton-devel list that you had confirmed
> that MLRISC no longer uses higher order modules.  I was wondering whether
> or not MLRISC uses any other "non-standard" Standard ML module features:
> e.g., functor definitions in local and struct blocks?  Also, are more
> current MLRISC source releases than the ones bundled with SML/NJ?  I note
> that that bundle still uses h.o. functors; and I've never had luck getting
> the demos (or even a majority of the .cm files) to compile.  Thanks.
> -Matthew
> .

This SF.NET email is sponsored by: FREE  SSL Guide from Thawte
are you planning your Web Server Security? Click here to get a FREE
Thawte SSL guide and find the answers to all your  SSL security issues.
MLton-devel mailing list