forbid(missing_docs) and nightly compiler

[ounsworth]

This PR stacks on top of #48

This PR is going to be a round-up of a bunch of docs changes and refactorings related to removing the nightly compiler requirement.

• Remove dependence on nightly / unstable
• All crates have #![forbid(missing_docs)] (and missing docs are added)

As a side-effect of removing nightly / unstable, I refactored how OIDs are carried for HashMLDSA to a more general trait in core, which actually is a much better design anyway.

[npajkovsky]

If you want to maintain your sanity at a reasonable level, try this

root/Cargo.toml

[workspace.lints]
rust.missing_docs = "warn"

and each crate will have in Cargo.toml

[lints]
workspace = true

[ounsworth]

I'll give that a try. Thanks.

[ounsworth]

I tried that with

[workspace.lints]
rust.missing_docs = "forbid"

in the workspace-level Cargo.toml, and

[lints]
workspace = true

in the crate-level Cargo.toml's, and yes that has the same effect, but I'm not a big fan because
A) as you're developing or debugging a crate, you might want to turn off specific lints and then turn them back on later, and
B) we may not want the same lints on every crates; for example I currently don't have forbid(missing_docs) on the cli or mem_usage_benches crates since those are not exposed as library crates.

So thank you for the suggestion, but I think I'll leave it the way it is.

[ounsworth]

@officialfrancismendoza I had to comment out his test at keccak.rs:545 because it's failing and I can't figure out what it's supposed to be testing, so I'll need you to debug it.

[ounsworth]

This PR stacks on top of #48

This PR is going to be a round-up of a bunch of docs changes and refactorings related to removing the nightly compiler requirement.

• Remove dependence on nightly / unstable
• All crates have #![forbid(missing_docs)] (and missing docs are added)

As a side-effect of removing nightly / unstable, I refactored how OIDs are carried for HashMLDSA to a more general trait in core, which actually is a much better design anyway.

[npajkovsky]

If you want to maintain your sanity at a reasonable level, try this

root/Cargo.toml

[workspace.lints]
rust.missing_docs = "warn"

and each crate will have in Cargo.toml

[lints]
workspace = true

[ounsworth]

I'll give that a try. Thanks.

[ounsworth]

I tried that with

[workspace.lints]
rust.missing_docs = "forbid"

in the workspace-level Cargo.toml, and

[lints]
workspace = true

in the crate-level Cargo.toml's, and yes that has the same effect, but I'm not a big fan because
A) as you're developing or debugging a crate, you might want to turn off specific lints and then turn them back on later, and
B) we may not want the same lints on every crates; for example I currently don't have forbid(missing_docs) on the cli or mem_usage_benches crates since those are not exposed as library crates.

So thank you for the suggestion, but I think I'll leave it the way it is.

[ounsworth]

@officialfrancismendoza I had to comment out his test at keccak.rs:545 because it's failing and I can't figure out what it's supposed to be testing, so I'll need you to debug it.

[ounsworth]

This PR stacks on top of #48

This PR is going to be a round-up of a bunch of docs changes and refactorings related to removing the nightly compiler requirement.

• Remove dependence on nightly / unstable
• All crates have #![forbid(missing_docs)] (and missing docs are added)

As a side-effect of removing nightly / unstable, I refactored how OIDs are carried for HashMLDSA to a more general trait in core, which actually is a much better design anyway.

[npajkovsky]

If you want to maintain your sanity at a reasonable level, try this

root/Cargo.toml

[workspace.lints]
rust.missing_docs = "warn"

and each crate will have in Cargo.toml

[lints]
workspace = true

[ounsworth]

I'll give that a try. Thanks.

[ounsworth]

I tried that with

[workspace.lints]
rust.missing_docs = "forbid"

in the workspace-level Cargo.toml, and

[lints]
workspace = true

in the crate-level Cargo.toml's, and yes that has the same effect, but I'm not a big fan because
A) as you're developing or debugging a crate, you might want to turn off specific lints and then turn them back on later, and
B) we may not want the same lints on every crates; for example I currently don't have forbid(missing_docs) on the cli or mem_usage_benches crates since those are not exposed as library crates.

So thank you for the suggestion, but I think I'll leave it the way it is.

[ounsworth]

@officialfrancismendoza I had to comment out his test at keccak.rs:545 because it's failing and I can't figure out what it's supposed to be testing, so I'll need you to debug it.

[ounsworth]

This PR stacks on top of #48

This PR is going to be a round-up of a bunch of docs changes and refactorings related to removing the nightly compiler requirement.

• Remove dependence on nightly / unstable
• All crates have #![forbid(missing_docs)] (and missing docs are added)

As a side-effect of removing nightly / unstable, I refactored how OIDs are carried for HashMLDSA to a more general trait in core, which actually is a much better design anyway.

[npajkovsky]

If you want to maintain your sanity at a reasonable level, try this

root/Cargo.toml

[workspace.lints]
rust.missing_docs = "warn"

and each crate will have in Cargo.toml

[lints]
workspace = true

[ounsworth]

I'll give that a try. Thanks.

[ounsworth]

I tried that with

[workspace.lints]
rust.missing_docs = "forbid"

in the workspace-level Cargo.toml, and

[lints]
workspace = true

in the crate-level Cargo.toml's, and yes that has the same effect, but I'm not a big fan because
A) as you're developing or debugging a crate, you might want to turn off specific lints and then turn them back on later, and
B) we may not want the same lints on every crates; for example I currently don't have forbid(missing_docs) on the cli or mem_usage_benches crates since those are not exposed as library crates.

So thank you for the suggestion, but I think I'll leave it the way it is.

[ounsworth]

@officialfrancismendoza I had to comment out his test at keccak.rs:545 because it's failing and I can't figure out what it's supposed to be testing, so I'll need you to debug it.

[ounsworth]

This PR stacks on top of #48

This PR is going to be a round-up of a bunch of docs changes and refactorings related to removing the nightly compiler requirement.

• Remove dependence on nightly / unstable
• All crates have #![forbid(missing_docs)] (and missing docs are added)

As a side-effect of removing nightly / unstable, I refactored how OIDs are carried for HashMLDSA to a more general trait in core, which actually is a much better design anyway.

[npajkovsky]

If you want to maintain your sanity at a reasonable level, try this

root/Cargo.toml

[workspace.lints]
rust.missing_docs = "warn"

and each crate will have in Cargo.toml

[lints]
workspace = true

[ounsworth]

I'll give that a try. Thanks.

[ounsworth]

I tried that with

[workspace.lints]
rust.missing_docs = "forbid"

in the workspace-level Cargo.toml, and

[lints]
workspace = true

in the crate-level Cargo.toml's, and yes that has the same effect, but I'm not a big fan because
A) as you're developing or debugging a crate, you might want to turn off specific lints and then turn them back on later, and
B) we may not want the same lints on every crates; for example I currently don't have forbid(missing_docs) on the cli or mem_usage_benches crates since those are not exposed as library crates.

So thank you for the suggestion, but I think I'll leave it the way it is.

[ounsworth]

@officialfrancismendoza I had to comment out his test at keccak.rs:545 because it's failing and I can't figure out what it's supposed to be testing, so I'll need you to debug it.

[ounsworth]

This PR stacks on top of #48

This PR is going to be a round-up of a bunch of docs changes and refactorings related to removing the nightly compiler requirement.

• Remove dependence on nightly / unstable
• All crates have #![forbid(missing_docs)] (and missing docs are added)

As a side-effect of removing nightly / unstable, I refactored how OIDs are carried for HashMLDSA to a more general trait in core, which actually is a much better design anyway.

[npajkovsky]

If you want to maintain your sanity at a reasonable level, try this

root/Cargo.toml

[workspace.lints]
rust.missing_docs = "warn"

and each crate will have in Cargo.toml

[lints]
workspace = true

[ounsworth]

I'll give that a try. Thanks.

[ounsworth]

I tried that with

[workspace.lints]
rust.missing_docs = "forbid"

in the workspace-level Cargo.toml, and

[lints]
workspace = true

in the crate-level Cargo.toml's, and yes that has the same effect, but I'm not a big fan because
A) as you're developing or debugging a crate, you might want to turn off specific lints and then turn them back on later, and
B) we may not want the same lints on every crates; for example I currently don't have forbid(missing_docs) on the cli or mem_usage_benches crates since those are not exposed as library crates.

So thank you for the suggestion, but I think I'll leave it the way it is.

[ounsworth]

@officialfrancismendoza I had to comment out his test at keccak.rs:545 because it's failing and I can't figure out what it's supposed to be testing, so I'll need you to debug it.

[ounsworth]

This PR stacks on top of #48

This PR is going to be a round-up of a bunch of docs changes and refactorings related to removing the nightly compiler requirement.

• Remove dependence on nightly / unstable
• All crates have #![forbid(missing_docs)] (and missing docs are added)

As a side-effect of removing nightly / unstable, I refactored how OIDs are carried for HashMLDSA to a more general trait in core, which actually is a much better design anyway.

[npajkovsky]

If you want to maintain your sanity at a reasonable level, try this

root/Cargo.toml

[workspace.lints]
rust.missing_docs = "warn"

and each crate will have in Cargo.toml

[lints]
workspace = true

[ounsworth]

I'll give that a try. Thanks.

[ounsworth]

I tried that with

[workspace.lints]
rust.missing_docs = "forbid"

in the workspace-level Cargo.toml, and

[lints]
workspace = true

in the crate-level Cargo.toml's, and yes that has the same effect, but I'm not a big fan because
A) as you're developing or debugging a crate, you might want to turn off specific lints and then turn them back on later, and
B) we may not want the same lints on every crates; for example I currently don't have forbid(missing_docs) on the cli or mem_usage_benches crates since those are not exposed as library crates.

So thank you for the suggestion, but I think I'll leave it the way it is.

[ounsworth]

@officialfrancismendoza I had to comment out his test at keccak.rs:545 because it's failing and I can't figure out what it's supposed to be testing, so I'll need you to debug it.

[ounsworth]

This PR stacks on top of #48

This PR is going to be a round-up of a bunch of docs changes and refactorings related to removing the nightly compiler requirement.

• Remove dependence on nightly / unstable
• All crates have #![forbid(missing_docs)] (and missing docs are added)

As a side-effect of removing nightly / unstable, I refactored how OIDs are carried for HashMLDSA to a more general trait in core, which actually is a much better design anyway.

[npajkovsky]

If you want to maintain your sanity at a reasonable level, try this

root/Cargo.toml

[workspace.lints]
rust.missing_docs = "warn"

and each crate will have in Cargo.toml

[lints]
workspace = true

[ounsworth]

I'll give that a try. Thanks.

[ounsworth]

I tried that with

[workspace.lints]
rust.missing_docs = "forbid"

in the workspace-level Cargo.toml, and

[lints]
workspace = true

in the crate-level Cargo.toml's, and yes that has the same effect, but I'm not a big fan because
A) as you're developing or debugging a crate, you might want to turn off specific lints and then turn them back on later, and
B) we may not want the same lints on every crates; for example I currently don't have forbid(missing_docs) on the cli or mem_usage_benches crates since those are not exposed as library crates.

So thank you for the suggestion, but I think I'll leave it the way it is.

[ounsworth]

@officialfrancismendoza I had to comment out his test at keccak.rs:545 because it's failing and I can't figure out what it's supposed to be testing, so I'll need you to debug it.