RFC: add revision format example for Buckets

Signed-off-by: Hidde Beydals <hello@hidde.co>
pull/3233/head
Hidde Beydals 2 years ago
parent fa217b8775
commit 2c12385344

@ -9,7 +9,7 @@ Must be one of `provisional`, `implementable`, `implemented`, `deferred`, `rejec
**Creation date:** 2022-10-20
**Last update:** 2022-10-20
**Last update:** 2022-10-21
## Summary
@ -17,8 +17,8 @@ This RFC proposes to establish a canonical format for an `Artifact` which
points to a specific checksum (e.g. an OCI manifest digest or Git commit SHA)
of a named pointer (e.g. an OCI image tag or Git tag). In addition, it proposes
to include the algorithm name (e.g. `sha256`) as a prefix to any advertised
checksum in an `Artifact` and further referring to it as a `Digest` opposed to
a `Checksum`.
checksum in an `Artifact` and further referring to it as a `Digest` as opposed
to a `Checksum`.
## Motivation
@ -67,7 +67,7 @@ with supportive response from Core Maintainers.
checksums. This has promising performance improvements over SHA-256, which
could allow for performance improvements in large scale environments.
- Allow compatability with SemVer name references which might contain an `@`
symbol already (e.g. `package@v1.0.0@sha256:...`, opposed to OCI's
symbol already (e.g. `package@v1.0.0@sha256:...`, as opposed to OCI's
`tag:v1.0.0@sha256:...`).
### Non-Goals
@ -108,6 +108,13 @@ without a defined branch or tag, the `Revision` field value would become:
sha1:e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855
```
For a `Bucket`'s `Artifact` with a revision based on an SHA-256 calculation of
a list of object keys and their etags, the `Revision` field value would become:
```text
sha256:e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855
```
### Change the Artifact Checksum to a Digest
Change the format of the `Checksum` field of the `source.toolkit.fluxcd.io`

Loading…
Cancel
Save