Re: MIT-Scheme (SVM1 version) on NetBSD/arm64

[Chris Hanson]

SVM is very slow; I’ve done informal measurements and found that it’s about 20x slower than native code.

On Jan 3, 2020, 4:50 PM -0500, Alexander Shendi (Web.DE) <address@hidden>, wrote:

Dear MIT-Scheme users,

I have managed to build a version of MIT-Scheme (V 10.1.10, SVM1) under NetBSD/aarch64 (aka arm64).

The resulting binary seems to run w/o problems, but when I try to start Edwin (either via "scheme --edit" or via the "(edwin)" procedure it takes a long time to initialize and show the X11 editor window. The Edwin editor then runs as expected.

Can somebody explain the delay and how to possibly avoid it?

Many thanks in advance,

Alexander
--
You have zero privacy anyway. Get over it.

Scott McNealy 1999