Skip to content

v3.0 release notes: contributor list appears to be generated at v3.0-rc3, three contributors missing #3989

Description

@h1219-kim

The contributor list in the v3.0 release notes looks like it was generated against v3.0-rc3 rather than the final v3.0 tag. Three people whose first mainline commits landed between rc3 and the release are missing from the list of 58.

Reproduction

Using the mainline-only walk the release notes describe:

$ git log --first-parent --format='%an' v2.16..v3.0-rc3 | sort -u > rc3.txt
$ git log --first-parent --format='%an' v2.16..v3.0     | sort -u > v30.txt
$ wc -l < rc3.txt
60
$ wc -l < v30.txt
63
$ comm -13 rc3.txt v30.txt
Hyuntae Kim
kanishka
mittal-ishaan

None of these three are in the release notes. All three also have zero commits before v2.16 — they are first-time contributors to the project, so unlike everyone else who landed after rc3, they had no earlier work in the cycle to put them on the list already. That is why the cutoff is only visible for these three.

The missing commits

Hyuntae Kim
  c8ec7e41f  tests: add CLI tests for samsung vs-internal-log
  b2e3069af  plugins/samsung: add Samsung vendor specific extensions plugin
  00e9818e4  shared: add shr_dir_prefix_len()
  8ac4c7995  shared: handle a path with no file name in shr_mkdir_from_fname()

kanishka
  bdb6e323a  tests: use pathlib in the e2e test cases
  57df8931c  tests: use pathlib in the e2e runner

mittal-ishaan
  03a67c2a2  libnvme: return an empty string for missing namespace attributes

Two side notes

  • Thomas Glanzmann has mainline commits in v2.16..v3.0-rc3 but is not on the list either. That one predates rc3, so it looks like a separate omission rather than part of the same cutoff.
  • The "1,711 commits" figure also sits in the rc3 range rather than the final one: --first-parent v2.16..v3.0-rc3 gives 1724 and v2.16..v3.0 gives 1795. I could not reproduce 1711 exactly with any filter I tried, so this is only weak corroboration, not a claim about the exact method used.

Disclosure: I am one of the three, and 3.0 is my first contribution to nvme-cli — which is exactly why the rc3 cutoff caught me. Raising it because the omission looks systematic rather than a one-off, and because the release notes are editable. No urgency whatsoever; 3.0 was a good release to land in, and the write-up was a genuinely good read.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions