Uhm build fails?

[MrFaul]

Uhm build fails?

[pitdicker]

Fixed with 0.4.33 in #1383. Thank you for reporting, and @jtmoon79 for investigating.

[jtmoon79]

I could not find failed builds here on github under the Actions tab. But I did find this build failure at docs.rs

...
[INFO] [stderr] warning: Target filter specified, but no targets matched. This is a no-op
[INFO] [stderr]     Checking rkyv v0.7.43
[INFO] [stderr] error: one of ["size_16", "size_32", or "size_64"] features must be enabled
[INFO] [stderr]    --> /opt/rustwide/cargo-home/registry/src/index.crates.io-6f17d22bba15001f/rkyv-0.7.43/src/macros.rs:114:1
[INFO] [stderr]     |
[INFO] [stderr] 114 | core::compile_error!(r#"one of ["size_16", "size_32", or "size_64"] features must be enabled"#);
[INFO] [stderr]     | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[INFO] [stderr] 
[INFO] [stderr] error: cannot find macro `pick_size_type` in this scope
[INFO] [stderr]    --> /opt/rustwide/cargo-home/registry/src/index.crates.io-6f17d22bba15001f/rkyv-0.7.43/src/lib.rs:723:23
...

At first glance, this looks to be an rkyv oddity.
Tag #644 and @dovahcrow

[jtmoon79]

Thanks @dovahcrow for following up too.

[MrFaul]

Thanks for fixing.

A small issue remains so:

All tests and builds are still failing at least according to the badge...
But I think there is something wrong with job naming and not the testing it self.

Since chrono is a very popular crate I think this isn't something that should left that way.
Even removing the badge is a better option.

[djc]

That's from the main branch, whereas we mostly release from the 0.4.x branch. The badge should fix itself when we do our next merge from 0.4.x to main. Perhaps we should rename main to 0.5.x and keep 0.4.x as main?

[dovahcrow]

I failed to reproduce locally. Here's what I ran:
cargo +nightly rustdoc --lib -Zrustdoc-map --features rkyv.
Rust version: v1.77.0-nightly

Seems to me --no-default-features is passed to rkyv somehow.

[MrFaul]

Would be better if the "landing page" is actually the current published version.
After all the same badges are also shown in the docs.

[pitdicker]

Perhaps we should rename main to 0.5.x and keep 0.4.x as main?

I'm all for it. We would have to change the target branch of all open PRs though.

[pitdicker]

I can't find the reason why the badge is still red but made an attempt in #1388.

Edit: still red because the latest merge is still running on the CI 😄.

[pitdicker]

Perhaps we should rename main to 0.5.x and keep 0.4.x as main?

I'm all for it. We would have to change the target branch of all open PRs though.

@djc I don't have permission, but it may just work when you rename the branches via the GitHub interface.

[djc]

Done -- let's use #1390 to coordinate any necessary changes in the repo contents.

[pitdicker]

Should we yank 0.4.32?

[djc]

Nah, it doesn't seem bad enough for that.

[MrFaul]

Nice 👍🏼
Looooks way better this way.
Thx for fixing it.