I hope Zig can stabilize as soon as possible. Copilot's auto-completion cannot even compile a HelloWorld (due to I/O-related changes), which is clearly not a good thing in the AI era. Time is already running out for new languages.
> Zig 0.16.0 is tagged. Release notes are just about completed and will be published in about 20 hours when the tarballs finish building.
As it turns out, it takes a while to build LLVM 27 times on one computer! That's how many different targets we provide compiler binaries for now - all cross compiled from one host.
I just saw in their CI pipeline that the build for x86_64 Linux failed. Their error however is different from the one occurring on my Linux machine, where the build complains when it is almost done that the put method needs 2 and not 3 parameters.
What does it all mean Andrew... do we have to wait longer now. THE AGONY.... Oh captain, my captain... :D
Hm, are you referring to the CI failure on the x86_64-linux-release CI job on the current tip of master on the main Zig repository? That one's a flaky test, not indicative of a serious problem. (We usually just disable flaky tests, but this particular one I'm leaving on for the moment to try and gather a little more information about it.)
I am so happy to here that thanks :) Ignore my complaint about the failing Zig build. I am using a script that also builds ZLS afterwards and it was ZLS that was failing not Zig. So sorry.
Men, the zig toolchain is just my goto c compiler. ty Andrew!
I hope Zig can stabilize as soon as possible. Copilot's auto-completion cannot even compile a HelloWorld (due to I/O-related changes), which is clearly not a good thing in the AI era. Time is already running out for new languages.
From Andrew’s post
> Zig 0.16.0 is tagged. Release notes are just about completed and will be published in about 20 hours when the tarballs finish building. As it turns out, it takes a while to build LLVM 27 times on one computer! That's how many different targets we provide compiler binaries for now - all cross compiled from one host.
I don’t use zig daily, I rather use Rust, but damn this is awesome. I wish Rust would do this too.
Do what? They do provide a significant number of prebuilt binaries?
I just saw in their CI pipeline that the build for x86_64 Linux failed. Their error however is different from the one occurring on my Linux machine, where the build complains when it is almost done that the put method needs 2 and not 3 parameters.
What does it all mean Andrew... do we have to wait longer now. THE AGONY.... Oh captain, my captain... :D
> the build for x86_64 Linux failed.
Hm, are you referring to the CI failure on the x86_64-linux-release CI job on the current tip of master on the main Zig repository? That one's a flaky test, not indicative of a serious problem. (We usually just disable flaky tests, but this particular one I'm leaving on for the moment to try and gather a little more information about it.)
The tarball builds for the 0.16.0 release are still moving along okay: https://codeberg.org/ziglang/www.ziglang.org/actions/runs/20...
I am so happy to here that thanks :) Ignore my complaint about the failing Zig build. I am using a script that also builds ZLS afterwards and it was ZLS that was failing not Zig. So sorry.