signature crate: 1.0 stabilization proposal #32
Conversation
|
@newpavlov on a semi-related issue, I'm wondering if we should move the |
| @@ -1,7 +1,7 @@ | |||
| [package] | |||
| name = "signature" | |||
| description = "Traits for cryptographic signature algorithms (e.g. ECDSA, Ed25519)" | |||
| version = "0.3.0" # Also update html_root_url in lib.rs when bumping this | |||
| version = "1.0.0-pre" # Also update html_root_url in lib.rs when bumping this | |||
There was a problem hiding this comment.
Note: if we do this, I'd like to do a series of prereleases (e.g. 1.0.0-pre.0, 1.0.0-pre.1) to work through issues encountered with ed25519-dalek, signatory, and potentially rsa as well /cc @roblabla
|
One thing I'd definitely like to do before a 1.0 release is improve the state of the documentation, both for consumers of the trait, and for crates that impl the provided traits. Per #25 there are a number of undocumented design decisions which need to be adequately described. |
tarcieri
force-pushed
the
signature-crate/1.0-stabilization-proposal
branch
from
5332de2 to
26553c1
Compare
The motivation for 1.0 stabilization of the `signature` crate is the upcoming 1.0 release of `ed25519-dalek`. In order to promote Ed25519 interoperability, it would be great if `ed25519-dalek` could use the traits from this crate along with the `ed25519::Signature` type from the `ed25519` crate. To get there, I think we need to do a 1.0 release of this crate, as well as the `ed25519` crate. The main impediment towards doing so is the `digest` crate is presently stuck at v0.8. It would be nice to be able to continue upgrading it, especially to a 1.0 release, but that would otherwise be a semver breaking change. To allow agility around `digest`, and `signatory_derive` which depends on it, this commit places access to both under the `digest-preview` and `derive-preview` Cargo features respectively, and calls them out as not covered under SemVer and subject to change, but breaking changes will be done with a minor version bump.
tarcieri
force-pushed
the
signature-crate/1.0-stabilization-proposal
branch
from
26553c1 to
b229861
Compare
|
Going to move forward with an initial 1.0 prerelease using this plan. If anyone would like to get any changes in prior to a final 1.0, I don't plan on being too hasty in cutting that. |
|
Note: final comment period for shipping |
The motivation for 1.0 stabilization of the
signaturecrate is the upcoming 1.0 release ofed25519-dalek. In order to promote Ed25519 interoperability, it would be great ifed25519-dalekcould use the traits from this crate along with theed25519::Signaturetype from theed25519crate. See the relevant issue:dalek-cryptography/ed25519-dalek#80
To get there, I think we need to do a 1.0 release of this crate, as well as the
ed25519crate.The main impediment towards doing so is the
digestcrate is presently stuck at v0.8. It would be nice to be able to continue upgrading it, especially to a 1.0 release, but that would otherwise be a SemVerbreaking change.
To allow agility around
digest, andsignatory_derivewhich depends on it, this commit places access to both under thedigest-previewandderive-previewCargo features respectively, and calls them out as not covered under SemVer and subject to change, but breaking changes will be done with a minor version bump.