This is something that really confused me when i first read some gleam code because it does not really look like a function call in other languages, which would be read from right to left but this way seems more logical when considering we read left to right and i think i would really get used to it when i get the opportunity to use it more
A general purpose memory allocator although this is really much a work in progress i think there are some good opportunities for otimization in a memory allocator for rust.
For example Rust gives you the size of memory region to free, which means the allocator does not have to track that.
Hmm now it would be interesting how eyra fares for allocating. And also why does musl not implement a faster allocator? I get that it should be backwards compatible but the gap to glibc seems to be really large.
And then you find out you have that dependency but your linker decides to not take it and then you have it but a slightly other version and you decide it's not worth it
This is my first post on the Threadiverse! I hope i didn't miss tagging the post with a flair. If you have any feedback feel free to write in this thread!
This is a smaller blogpost but i will write larger ones in the future hopefully more in-depth than those in the past
This is something that really confused me when i first read some gleam code because it does not really look like a function call in other languages, which would be read from right to left but this way seems more logical when considering we read left to right and i think i would really get used to it when i get the opportunity to use it more
A general purpose memory allocator although this is really much a work in progress i think there are some good opportunities for otimization in a memory allocator for rust.
For example Rust gives you the size of memory region to free, which means the allocator does not have to track that.
Maybe Cloudflare's pingora suits this purpose better given that it is being used by cloudflare to proxy traffic at a large scale
Yeah that just wastes both people's time
Yeah that's the fun part!
Maybe there are also some security implications of the code?
Because the thing is: That code is probably gonna end up in production somewhere
Hmm now it would be interesting how eyra fares for allocating. And also why does musl not implement a faster allocator? I get that it should be backwards compatible but the gap to glibc seems to be really large.
And then you find out you have that dependency but your linker decides to not take it and then you have it but a slightly other version and you decide it's not worth it
This is my first post on the Threadiverse! I hope i didn't miss tagging the post with a flair. If you have any feedback feel free to write in this thread!
This is a smaller blogpost but i will write larger ones in the future hopefully more in-depth than those in the past