Bump bitcoin v0.29.1

[DanGould]

depends on Kixunil/bip21#8

[DanGould]

I'm not sure why this is failing CI on edition key. All the updates were to projects with MSRV 1.41.1? @Kixunil may you help me out?

[Kixunil]

The binary client has higher MSRV than 1.41.1 and it depends on other things that were updated. Just revert those updates and we can decide later if we need to do something about them.

[Kixunil]

Apart from build failures this looks OK.

[Kixunil]

Oh, we forgot IntoIterator for &'_ Witness? Damn!

Made a PR right away: rust-bitcoin/rust-bitcoin#1354

[DanGould]

I think there's something deeper wrong with the build here. Even if I revert, there's an old dependency on core-rpc which was yanked from crates.rs. So a reverted build won't pass CI any longer.

Do you know exactly what dependency is msrv noncompliant? I looked at all of their docs and they seem to be. The tests failed at "once_cell" which is a build dependency of bitcoind v0.27.1. v0.27.* claim to be MSRV 0.41. Am I reading the output of cargo tree wrong?

[Kixunil]

I think it's MSRV 1.41.1 but only if you use older once_cell and hashbrown. @RCasatta do you happen to know which versions work? I don't have much time to investigate rn.

[DanGould]

I filed an issue in bitcoind which accidentally breaks its 1.41.1 MSRV promise because of a downstream package.

Is there a good reason we're supporting all the way back to 1.41 on this crate? This seems like more of an application-layer crate that would sit at roughly the same level of software as bdk, which is 1.56 MSRV.

[Kixunil]

I'd like to move this into the rust-bitcoin org eventually which requires 1.41.1 and for my projects I require whatever is in Debian stable - currently 1.48. But we're considering bumping to 1.48 in rust-bitcoin too so that wouldn't be too bad.

The failures are for some tests only though so I wouldn't mind part of the tests being ran on higher MSRV. (Not sure what the people at rust-bitcoin would think about that.)