Skip to content

openkal 0.11 and the eight packages that move with it - #315

Merged
Sunrisepeak merged 1 commit into
mainfrom
openkal-0.11-ecosystem
Aug 30, 2026
Merged

openkal 0.11 and the eight packages that move with it#315
Sunrisepeak merged 1 commit into
mainfrom
openkal-0.11-ecosystem

Conversation

@Sunrisepeak

Copy link
Copy Markdown
Member

⚠️ THIS REVISION REMOVES DECLARATIONS, which every earlier one was careful not
to. kal_process_spawn_with' and kal_process_spawn_bound' are gone: they were
never three operations, they were one operation and three modifiers, and the
combinations were about to become sixteen. Code written against them fails to
COMPILE. Clause 8 was set aside deliberately and once, at the last version where
there is no written code to protect.

  • openkal 0.11.0 --- one spawn taking a kal_spawn' record; kal_job' as a
    resource the caller holds; `kal_process_stop_requested'
  • openkal-linux 0.10.0, openkal-macos 0.8.0, openkal-windows 0.6.0 --- each
    implements what its environment can, and declines the rest by not claiming the
    position rather than by approximating it
  • openkal-opensbi 0.5.0, openkal-uefi 0.5.0 --- repins
  • openkal-musl 0.12.0 --- takes it up: a started program runs where the caller
    is, setpgid'/kill(-n)' form and end a real unit, and the `execve' waiter
    stops holding a pipe open on behalf of a program that has ended
  • openkal-llvm-runtime 0.8.0 --- carries it

⭐ THE UNIT'S SHAPE WAS DECIDED BY CLAUSE 7.1 AND NOT BY PREFERENCE. As a flag it
needed a registry on one implementation, and clause 7.1 states mechanically that
an implementation needing a registry indicates the SPECIFICATION borrowed a
shape. The obvious repair --- an operation opening an empty unit --- would have
moved the same defect to the other kind of system, so the identity is established
at the first start and reported back, which both perform remembering nothing.

Every sha256 is of the tarball GitHub serves, fetched when the entry was written,
and the byte-identical file was uploaded to gitcode and READ BACK AND COMPARED.
⚠️ openkal-musl's first upload failed and was retried and re-verified; an entry
names one hash for both URLs, so an unverified CN mirror would fail for every CN
user and for nobody else.

std-freestanding-alloc-kal 0.1.4 is NOT here and cannot be: its CI resolves
openkal from the PUBLISHED index artifact, which republishes only when this
lands. Same ordering as the 0.10 round.

⚠️ THIS REVISION REMOVES DECLARATIONS, which every earlier one was careful not
to. `kal_process_spawn_with' and `kal_process_spawn_bound' are gone: they were
never three operations, they were one operation and three modifiers, and the
combinations were about to become sixteen. Code written against them fails to
COMPILE. Clause 8 was set aside deliberately and once, at the last version where
there is no written code to protect.

- openkal 0.11.0 --- one spawn taking a `kal_spawn' record; `kal_job' as a
  resource the caller holds; `kal_process_stop_requested'
- openkal-linux 0.10.0, openkal-macos 0.8.0, openkal-windows 0.6.0 --- each
  implements what its environment can, and declines the rest by not claiming the
  position rather than by approximating it
- openkal-opensbi 0.5.0, openkal-uefi 0.5.0 --- repins
- openkal-musl 0.12.0 --- takes it up: a started program runs where the caller
  is, `setpgid'/`kill(-n)' form and end a real unit, and the `execve' waiter
  stops holding a pipe open on behalf of a program that has ended
- openkal-llvm-runtime 0.8.0 --- carries it

⭐ THE UNIT'S SHAPE WAS DECIDED BY CLAUSE 7.1 AND NOT BY PREFERENCE. As a flag it
needed a registry on one implementation, and clause 7.1 states mechanically that
an implementation needing a registry indicates the SPECIFICATION borrowed a
shape. The obvious repair --- an operation opening an empty unit --- would have
moved the same defect to the other kind of system, so the identity is established
at the first start and reported back, which both perform remembering nothing.

Every sha256 is of the tarball GitHub serves, fetched when the entry was written,
and the byte-identical file was uploaded to gitcode and READ BACK AND COMPARED.
⚠️ openkal-musl's first upload failed and was retried and re-verified; an entry
names one hash for both URLs, so an unverified CN mirror would fail for every CN
user and for nobody else.

⇒ `std-freestanding-alloc-kal` 0.1.4 is NOT here and cannot be: its CI resolves
openkal from the PUBLISHED index artifact, which republishes only when this
lands. Same ordering as the 0.10 round.
@Sunrisepeak
Sunrisepeak merged commit 9a7f40d into main Aug 30, 2026
8 checks passed
@Sunrisepeak
Sunrisepeak deleted the openkal-0.11-ecosystem branch August 30, 2026 20:34
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant