Just wish it had a native version for Windows on Arm. I tried compiling it but failed during libfffi due to the build system choosing x86 assembly. Need some patches there, maybe the asm from the aarch64 Linux tree works.
The latest report talks about this in some detail. Some things are faster, some slower and the slow things are going to get better in time. Compilation times are longer.
One should not trivialize the amount of work that went into porting all of the macro shenanigans to Chez, and the amount of compatibility bridges that needed to done (see figure 3 in this document) to actually make it compatible, semantically.
Downloading it now.
Just wish it had a native version for Windows on Arm. I tried compiling it but failed during libfffi due to the build system choosing x86 assembly. Need some patches there, maybe the asm from the aarch64 Linux tree works.
Very cool! This blew my mind when I first heard about it.
What does this mean to racket? Will this make the executables fast or does it make the compilation fast?
The latest report talks about this in some detail. Some things are faster, some slower and the slow things are going to get better in time. Compilation times are longer.
Why? From what I’ve seen, they’ve done much more mind-blowing things than re-basing Racket on Chez?
One should not trivialize the amount of work that went into porting all of the macro shenanigans to Chez, and the amount of compatibility bridges that needed to done (see figure 3 in this document) to actually make it compatible, semantically.