30237489589
92%

Ran 27 Jul 2026 04:40AM UTC

Jobs 1

Files 129

Run time 1min

Badge

Embed ▾

Committed 27 Jul 2026 04:34AM UTC coverage: 91.887%. Remained the same

Build # 30237489589

Build Type

push

github

Committed by

steve-downey

Commit Message

Merge blog phase 21: Common Lisp control flow as sender graphs

The payoff post for the pivot's thesis: phase 15 argued that CL's
dynamic-extent control operators are what a structured-concurrency backend can
express soundly; this post shows it running.

The spine is that a sender's three completion channels are a semantic resource.
The closure backends multiplex a value, a diagnosed error and two kinds of
unwind onto one result<value>, discriminated by sentinel parse_error messages
compared by pointer identity. Mapping the same language onto senders deletes
that machinery rather than porting it, and turns unwind-protect into an
ordinary adapter whose three completion functions funnel into one cleanup --
making "cleanup on every exit path" structural instead of a checklist.

The post keeps the counterweight in view: dynamic-binding restore does not move
to the sender world, because it brackets a callee's whole extent rather than a
single sender's completion. A sender backend is better exactly where the thing
being expressed is a completion.

Seven of the ten available anchors, all pinned to blog/phase-21, which existed
before the prose was written. DRAFT pending author revision.

Coverage Stats

10363 of 11278 relevant lines covered (91.89%)

249.78 hits per line

Jobs

ID	Job ID	Ran	Files	Coverage
1	30237489589.1	27 Jul 2026 04:40AM UTC	129	91.89	GitHub Action Run

steve-downey / compile-time-scheme / 30237489589
92%

README BADGES
x

Markdown

Textile

RDoc

HTML

Rst

Jobs

Source Files on build 30237489589

steve-downey / compile-time-scheme / 30237489589 92%

README BADGES x

Markdown

Textile

RDoc

HTML

Rst

Jobs

Source Files on build 30237489589

steve-downey / compile-time-scheme / 30237489589
92%

README BADGES
x