feat!: update to edition 2024 and update deps to latest

[dignifiedquire]

Description

Updates dependencies where possible.

Closes #3385

Depends on

• feat!: edition2024, MSRV 1.85 and update deps to latest net-tools#30

Breaking Changes

All crates are now edition2024

[github-actions]

Documentation for this PR has been generated and is available at: https://n0-computer.github.io/iroh/pr/3386/docs/iroh/

Last updated: 2025-07-23T09:36:11Z

[github-actions]

Netsim report & logs for this PR have been generated and is available at: LOGS
This report will remain available for 3 days.

Last updated for commit: 44f53ca

[rklaehn]

If it works, it works, I guess.

But tbh I would prefer to use 'static wherever possible to indicate that the returned thing is fully self-contained.

Especially this one is probably redundant:

mpl Iterator<Item = hickory_resolver::proto::rr::Record> + 'static + use<>

It's an iterator that does not have a reference on any of the lifetimes of the input parameters. Why is the use<> needed then?

[flub]

Now I know the intention from the other PR: all the Cargo.toml files in this PR also don't bump the edition.

[rklaehn]

There are lots of places where we have fn foo(&self) -> impl Bar + use<>.

As far as I can see all these places can be just replaced with -> impl Bar + 'static, which for me reads much more natural than the weird use<> syntax.

[dignifiedquire]

As far as I can see all these places can be just replaced with -> impl Bar + 'static, which for me reads much more natural than the weird use<> syntax.

But asfaiu that does not mean the same thing though

[rklaehn]

As far as I can see all these places can be just replaced with -> impl Bar + 'static, which for me reads much more natural than the weird use<> syntax.

But asfaiu that does not mean the same thing though

Not in general, but in this case? The returned thing is supposed to have no lifetime dependencies on any of the parameters that are passed by reference (which is just &self). So how is that not the same as 'static? What other source of lifetimes other than the args is there?

[dignifiedquire]

+ 'static definitely does sth different, the compiler is quite upset if I try to use it

[rklaehn]

+ 'static definitely does sth different, the compiler is quite upset if I try to use it

If you just use 'static everywhere instead of use<> everything still compiles. #3392

I will have to try out in play.rust-lang.org what the rules are from calling use<> from 'static or otherwise.

[dignifiedquire]

I will have to try out in play.rust-lang.org what the rules are from calling use<> from 'static or otherwise.

asfaict use<> avoids capturing the lifetime of &self, which I think 'static does not do

[rklaehn]

I will have to try out in play.rust-lang.org what the rules are from calling use<> from 'static or otherwise.

asfaict use<> avoids capturing the lifetime of &self, which I think 'static does not do

'static does not capture &self. Static means fully self-contained, so you can use the future even after self goes out of scope. That is why tokio::task::spawn requires 'static. I really don't get what the difference is between use<> and 'static in that sense.

[matheus23]

We could merge this as-is probably, but we could follow this up with #3198 as well.

But we'd need another iroh-quinn release where we merge upstream's quinn-rs/quinn#2144