update deps

[ognevny]

fixes #5116
just for interest, is it good to update other dependencies?
(hope lock file update doesn't break anything. tested only on x86_64-pc-windows-gnu toolchain)

[ognevny]

oh, I see the CI results. will try to work on it later

[fanninpm]

The CI failures seem unrelated.

[ognevny]

oh, I see the CI results. will try to work on it later

should be fixed. tested on Linux machine

[ognevny]

the Redox failure is very strange

[ognevny]

now not all yanked deps are updated :(

[youknowone]

@ognevnydemon Oh, I am sorry. This patch was independently updating Cargo.lock, right? I will try to lookup previous version.

[ognevny]

@ognevnydemon Oh, I am sorry. This patch was independently updating Cargo.lock, right? I will try to lookup previous version.

yeah, but I just ran cargo update which is not correct in many cases... so maybe it's needed to update them carefully

[youknowone]

just for interest, is it good to update other dependencies?

Yes, we update dependencies for various reasons.
Without getting new features, these are good reasons to update dependencies.

• To remove one of multiple versions of same library in Cargo.lock
• To follow up incompatible library update. (major upgrade for >= 1.0.0 or minor upgrade for <1.0.0)

Avoiding yanked version is one of the good reason to do it. Any other reasonable suggestion will be also welcome.

[ognevny]

Yes, we update dependencies for various reasons.
Without getting new features, these are good reasons to update dependencies.

• To remove one of multiple versions of same library in Cargo.lock
• To follow up incompatible library update. (major upgrade for >= 1.0.0 or minor upgrade for <1.0.0)

Avoiding yanked version is one of the good reason to do it. Any other reasonable suggestion will be also welcome.

understood. thanks for attention! maybe I'll try to update more deps if they have more fresh version

[youknowone]

uh.. openssl has changes

[youknowone]

which crate would affect test_fractions?

[youknowone]

maybe num-complex or other num- one

[ognevny]

I rebased the changes, now this PR updates only lockfile

[youknowone]

The change is partially merged in #5217

[ognevny]

looks like only one test fails. is this critical?

[ognevny]

The change is partially merged in #5217

yes, I rebased my branch to main so now it's just cargo update change

[youknowone]

It is hard to say critical, but at the same time, updating deps is not also critical.
Merging without reasoning doesn't seem reasonable.
It will be worth to try revert only num-* crates and retrying it to check what's going on.