Cx Dev Log — 2026-05-30

#cx #programming

No developer commits landed today. The working tree on main is clean. The matrix holds at 182 PASS / 0 FAIL. This is the kind of day where nothing moves forward, but the gap between what exists on submain and what is on main keeps widening silently.

The integration debt picture

Submain is now 10 non-merge commits ahead of main. That backlog includes real work: lazy arena allocation (eliminating the 64 KB zero-fill per function call), BindingId-indexed variables, a builtin registry, array bounds checking, composite literal type-checking, the exit() builtin, CLI --help/--version flags, human-readable token names in parse errors, and a CRLF fix for the matrix harness on Windows. Most of this landed in two bursts on May 24-25 and May 29.

Fourteen daily-log branches are also unmerged. Each carries incremental roadmap corrections. The last daily-log branch that reached main was from May 9, over three weeks ago.

The merge itself should be straightforward. Submain merged main via PR #282 on May 24, so there are no expected conflicts. But the gap keeps growing, and every daily log since May 6 has predicted the merge as the top next step. This is the sixth consecutive day making that prediction.

Roadmap corrections

The roadmap on main still says "0.1 release candidate" and "117/117 matrix." Neither has been true for a while. The actual state is v0.1.0 released (the tag exists) and 182 tests passing.

Phase 11 (surface area reduction) is now fully complete. The three remaining sub-items were confirmed landed on main via their commit hashes:

Method call actual lowering (commit 0ab7e9b)
when block lowering (commit bed71c1)
DotAccess in compound forms

These had been sitting in the codebase for some time but were never checked off in the roadmap. The daily-log commits carry these corrections, but since none of those branches have merged to main, the roadmap file on main remains stale.

What is still waiting

Phase 8 Round 2 (str/strref layout, Handle) is still incomplete. Phases 12, 14, and 15 live on submain and haven't reached main. The remaining print intrinsics (print, println, printn, read, input) are still pending from Phase 9. JIT bounds checking (tracker #003) is the natural follow-up to the lazy arena and array bounds work from last week's sprint.

Looking ahead

The single highest-leverage action remains the same one it has been for weeks: merge submain to main. Ten commits carrying performance improvements, correctness fixes, CLI polish, and build improvements are waiting. The 14 pending daily-log branches could be batch-merged afterward to bring the roadmap file back in sync with reality.

Whether the next session breaks the pattern or adds a seventh day to the prediction streak remains to be seen.

Follow the Cx language project: