bootstrap: update some deps #92480
bootstrap: update some deps #92480
Conversation
|
(rust-highfive has picked a reviewer for you, use r? to override) |
rust-highfive
added
the
S-waiting-on-review
src/bootstrap/dist.rs
Outdated
| .unwrap_or_else(|_| time::now()); | ||
| .unwrap_or_else(|_| OffsetDateTime::now_utc()); |
There was a problem hiding this comment.
Used https://docs.rs/time/0.3.5/time/struct.OffsetDateTime.html#method.now_utc, but maybe https://docs.rs/time/0.3.5/time/struct.OffsetDateTime.html#method.now_local should be used instead?
There was a problem hiding this comment.
What did the previous function do? Let's try to match that.
There was a problem hiding this comment.
Looking at description at 0.1.x version, it says Returns the current time in the local timezone, so i guess it is.
From 0.3.6 (https://github.com/time-rs/time/blob/main/CHANGELOG.md#036-2022-01-20) now_local can be used again on more platforms.
|
☔ The latest upstream changes (presumably #92482) made this pull request unmergeable. Please resolve the merge conflicts. |
Mark-Simulacrum
added
S-waiting-on-author
|
Marking as waiting on author per #92480 (comment) -- would like to see a response there. It's worth noting that I suspect we could just drop the time crate dependency by dropping and/or hardcoding dates in man pages, which are likely not terribly important -- and seem to be the only use? |
joelpalmer
added
S-waiting-on-author
pretty_assertions 0.6 -> 0.7 to drop ansi_term 0.11 dependency update serde_json to dedupe itoa duplicate (from time update)
klensy
force-pushed
the
bootstrap-deps
branch
from
d1be5cb
to
d34c9a8
Compare
|
@rustbot ready Perhaps shouldn't be rolled up: may fail on some platforms (i hope this was fixed) plus serde_json may change perf? |
rustbot
added
S-waiting-on-review
|
I've filed #93685 to drop the time crate dependency entirely -- that should simplify this PR, and I'd like to wait on that as I'm not so enthusiastic about the num-threads library being pulled in. |
|
|
|
Given that #93685 will be merged, i'll close this one and open other one, as there some things that can be cared about, and not bootstrap specific. |
…albini Drop time dependency from bootstrap This was only used for the inclusion of 'current' dates into our manpages, but it is not clear that this is practically necessary. The manpage is essentially never updated, and so we can likely afford to keep a manual date in these files. It also seems possible to just omit it, but that may cause other tools trouble, so avoid doing that for now. This is largely done to reduce bootstrap complexity; the time crate is not particularly small and in rust-lang#92480 would have started pulling in num-threads, which does runtime thread count detection. I would prefer to avoid that, so filing this to just drop the nearly unused dependency entirely. r? `@pietroalbini`
…albini Drop time dependency from bootstrap This was only used for the inclusion of 'current' dates into our manpages, but it is not clear that this is practically necessary. The manpage is essentially never updated, and so we can likely afford to keep a manual date in these files. It also seems possible to just omit it, but that may cause other tools trouble, so avoid doing that for now. This is largely done to reduce bootstrap complexity; the time crate is not particularly small and in rust-lang#92480 would have started pulling in num-threads, which does runtime thread count detection. I would prefer to avoid that, so filing this to just drop the nearly unused dependency entirely. r? `@pietroalbini`
…ulacrum Drop time dependency from bootstrap This was only used for the inclusion of 'current' dates into our manpages, but it is not clear that this is practically necessary. The manpage is essentially never updated, and so we can likely afford to keep a manual date in these files. It also seems possible to just omit it, but that may cause other tools trouble, so avoid doing that for now. This is largely done to reduce bootstrap complexity; the time crate is not particularly small and in rust-lang#92480 would have started pulling in num-threads, which does runtime thread count detection. I would prefer to avoid that, so filing this to just drop the nearly unused dependency entirely. r? `@pietroalbini`
update time 0.1 -> 0.3 to mitigate RUSTSEC-2020-0071 (partial changelog: https://github.com/time-rs/time/blob/592ef915cf6275069f65483216ba7ccb8081b2bc/CHANGELOG.md )
Crate
timewill be duplicated until other crates will be updated (not all of them released patched versions)update pretty_assertions 0.6 -> 0.7 to drop ansi_term 0.11 dependency (changelog: https://github.com/colin-kiegel/rust-pretty-assertions/blob/da3cf9ad7a0a1120105efcd607b2b8c2bda37213/CHANGELOG.md)
update serde_json 1.0.59 -> 1.0.78 to dedupe
itoa(it was added intimeupdate). There no separate changelog for serde_json.