[plt-dev] Re: Plans for Unsafe f64vectors

From: Matthew Flatt (mflatt at cs.utah.edu)
Date: Tue Dec 15 16:24:08 EST 2009

We have not yet thought much beyond `f64vector' and `flvector', but I
expect that we'll get there eventually.

At Mon, 14 Dec 2009 12:15:33 -0700, Doug Williams wrote:
> Matthew,
> 
> I'm back looking into an efficient homogeneous array implementation to
> underlie a Matlab-like capability for PLT Scheme. My reference
> implementation sits on top of SRFI 4 (which I believe sits on to of
> scheme/foreign), but has rather poor performance. I've seen a couple
> references to unsafe f64vector operations on the mailing list and looked at
> them in the subversion build. Is there a long-term plan for performance
> improvements for homogeneous vectors? Can this be extended to include other
> homogeneous vector types (single floats and various sizes of signed and
> unsigned integers)? This could provide a sound and efficient basis for an
> efficient homogeneous array implementation.
> 
> Doug


Posted on the dev mailing list.