Roadmap

Open Issues

#160 — Concurrent plcc-scan/plcc-make invocations race on shared build dir Two CLI invocations sharing the same ./plcc-ng/ build dir race on temp-file creation/cleanup and crash with a raw FileNotFoundError traceback instead of a friendly error.
#186 — plcc-rep deadlocks on a partial stdout line A semantic action that writes without a trailing newline merges with the JSON result line, so _read_response destroys the result and blocks forever — exit 124, no stdout, no stderr.
#190 — A program error kills the plcc-rep session and blames the specification Any exception a semantic action raises that is not a LanguageError becomes a specification_error, so one bad input ends the session and tells the user to go fix a specification that is not at fault.
#191 — Interpreter resource exhaustion is reported as a specification error Exhausting the target runtime's call stack reads as Specification error: RecursionError, and the ceiling differs about eightfold between Python and the other targets, so the same program is fine on two of them.

#161 — Consider renaming plcc-rep to plcc-eval for phase-naming consistency plcc-rep is named after its interaction mode (REPL), not its phase, breaking the scan/parse/? naming pattern; an alias or rename to plcc-eval would restore it.
#187 — plcc-rep lacks output and clean-exit record kinds Semantic actions have no supported channel for user-visible output and no way to end the session cleanly, so output must be buffered into the result and a deliberate exit reads as a crash.

#185 — plcc-rep parses each SOURCE independently, unlike PLCC's rep Old PLCC's rep joined its file arguments into one stream; plcc-ng parses each separately, so a program split across two files no longer parses — an undocumented breaking change.

#184 — Follow-ups left open by the docs example tier Six non-blocking observations from the #181/#182/#183 reviews: the strict gate re-enumerates mkdocs.yml's plugin list, plus five diagnostics and test-hygiene items.

#154 — Update python-semantic-release Pinned to 9.x (locked 9.21.2); latest is 10.5.3. Dev-only dependency, consider updating.
#156 — Decide our MkDocs 1.x successor mkdocs-material hard-pins mkdocs<2; mkdocs-kroki-plugin already pulls in properdocs. Not urgent yet, but we'll need to pick ProperDocs, Zensical, or stay pinned once MkDocs 1.x actually breaks.
#189 — Adopt the languages-ng issue-system shape: issues never move, status is frontmatter Moving issues to done/ on close keeps breaking links (14 are broken right now, despite #150's rewriting); making closed: a frontmatter date instead deletes the bug class and the machinery guarding it.