Consider hint-mostly-unused rustc flag

[A6GibKm]

Since users might need only a couple of functions it might make sense to add the hint-mostly-unused flag.

See

• https://blog.rust-lang.org/inside-rust/2025/07/15/call-for-testing-hint-mostly-unused/
• https://doc.rust-lang.org/nightly/unstable-book/compiler-flags/hint-mostly-unused.html

for more information.

Of course this would need some accompanying numbers to justify it.

[A6GibKm]

Building lorem results in:

without this flag:

a 29.0s build time and:

with the flag

a 28.0s build time and:

The used commands where:

cargo +nightly build --release --timings
cargo +nightly clean
cargo +nightly -Zprofile-hint-mostly-unused build --release --timings

after adding

[profile.release.package.libadwaita]
hint-mostly-unused = true

[profile.release.package.gtk4]
hint-mostly-unused = true

[profile.release.package.glib]
hint-mostly-unused = true

[profile.release.package.gio]
hint-mostly-unused = true

[profile.release.package.pango]
hint-mostly-unused = true

[sdroege]

That doesn't seem very impressive, unfortunately :)

[A6GibKm]

It does not, hence why it would be interesting to get more feedback or see if my test is flawed somehow. The graphs above show better numbers than the total for some reason, but it does not lower the total much for some reason.