Make Cargo-minimal and Cargo-recent lockfiles follow expected convention

[benalleng]

The Cargo-minimal and Cargo-recent lockfiles were not actually pointing to the minimal and recent semvers of the dependencies and were instead just copies of the Cargo.lock lockfile.

This is somewhat related to the release but I don't think its all that critical just something that I thought should be improved.

Relates to #940 though I'm not sure it should close it
also related is #454

Pull Request Checklist

Please confirm the following before requesting review:

• I have disclosed my use of
  AI
  in the body of this PR.
• I have read CONTRIBUTING.md and rebased my branch to produce hygienic commits.

[coveralls]

Coverage Report for CI Build 27358652152

Coverage remained the same at 85.188%

Details

• Coverage remained the same as the base build.
• Patch coverage: No coverable lines changed in this PR.
• No coverage regressions found.

Uncovered Changes

No uncovered changes found.

Coverage Regressions

No coverage regressions found.

---

Coverage Stats

Relevant Lines:	14718
Covered Lines:	12538
Line Coverage:	85.19%
Coverage Strength:	371.0 hits per line

---

💛 - Coveralls

[benalleng]

This should avoid cargo updates from occuring on stray CI jobs preventing failures like https://github.com/payjoin/rust-payjoin/actions/runs/27285126244/job/80590095693?pr=1629 in efc5b92 from happening

[DanGould]

This is a welcome cleanup for a dangling issue

I ran into an issue running ./update-lock-files.sh locally where payjoin-test-utils had futures = "0.3" and payjoin-mailroom has futures = "0.3.21". cargo +nightly check -Z direct-minimal-versions pins each direct dep to
its own lowest acceptable version, picks 0.3.11 for test-utils, then hits mailroom's ≥0.3.21 floor — and since there's one futures node in the graph, the minima can't coexist → "all possible versions conflict."

error from run

```console dan@Excavator payjoin % bash ./contrib/update-lock-files.sh Updating crates.io index Updating git repository `https://github.com/chavic/uniffi-bindgen-cs.git` Updating git repository `https://github.com/Uniffi-Dart/uniffi-dart.git` error: failed to select a version for `futures`. ... required by package `payjoin-mailroom v0.1.1 (/Users/dan/f/dev/payjoin/payjoin-mailroom)` versions that meet the requirements `^0.3.21` are: 0.3.21

all possible versions conflict with previously selected packages.

previously selected package futures v0.3.11
... which satisfies dependency futures = "^0.3" of package payjoin-test-utils v0.0.1 (/Users/dan/f/dev/payjoin/payjoin-test-utils)
... which satisfies path dependency payjoin-test-utils of package payjoin v1.0.0-rc.3 (/Users/dan/f/dev/payjoin/payjoin)
... which satisfies dependency payjoin = "^1.0.0-rc.3" of package payjoin-cli v0.2.0 (/Users/dan/f/dev/payjoin/payjoin-cli)

failed to select a version for futures which could resolve this conflict

[DanGould]

Bumping this fixes my update run locally

Suggested change

	futures = { version = "0.3", default-features = false, features = ["std"] }
	futures = { version = "0.3.21", default-features = false, features = ["std"] }

[benalleng]

I have no idea how I didn't catch this, good find

[DanGould]

I think we want --all-targets here so that we make sure tests (and examples, if we have them) also compile with these minimum deps. I don't think dev-dependencies find their minimums without this.

Suggested change

	rm -f Cargo.lock && cargo +nightly check --all-features -Z direct-minimal-versions
	rm -f Cargo.lock && cargo +nightly check --all-features --all-targets -Z direct-minimal-versions

[DanGould]

Do you know why this was separated in the first place? Chesterson's fence?

[benalleng]

I looked into the history for this very question. I think that we simply didn't want any failures to hit the main crate at the time as there is no other explanation was given to keep them separate. #846 included commits to actually make it pass the lint so I suspect that was partly why as there was perhaps some expectation of continued instability. @arminsabouri do you have a recollection about any additional motivation here?

[DanGould]

I know the workspaces used to be separated because there was no way to compile payjoin-ffi with the old MSRV, so that'd be one potential reason. Now everything is in the top level Cargo.toml workspace it can all be compiled and linted together.

[DanGould]

My only real concern here is 'Merge lint and lint-ffi into a single CI job' being here without rationale or understanding why they were separated in the first place. Because it's linting in CI I don't think it's load bearing and am ok moving forward but I do wonder why