Document native JavaScript syntax for eDSLs - #1495
Conversation
52e3e67 to
023c734
Compare
Deploying with
|
| Status | Name | Latest Commit | Preview URL | Updated (UTC) |
|---|---|---|---|---|
| ✅ Deployment successful! View logs |
functionalscript | 023c734 | Commit Preview URL Branch Preview URL |
Aug 12 2026, 02:36 AM |
o2alexanderfedin
left a comment
o2alexanderfedin
left a comment
There was a problem hiding this comment.
Reviewed at 023c7347dd707e097e05ae0ae754f2a7c73b424c.
This and #1496 are the same change written twice — both insert ### 5.8 at the same
point in AGENTS.md from the same pre-image (f4164b61d), so they conflict by
construction and only one should land. The text here is accurate and §8.3 does exempt
AGENTS.md from a CHANGELOG entry, so this is purely a question of which draft to keep.
I put the comparison on #1496, which is the one I'd keep — it carries the RTTI and NaNVM
evidence in the added text rather than only in the PR body, and it states which layer the
public API should be optimized for rather than only permitting the two layers to differ.
The one substantive thing this version does that #1496 doesn't is name existing eDSLs —
"HTML, BNF, RTTI, generated-code/test descriptions" — as targets for the principle. That
cuts both ways: BNF's tagged encoding is deliberate and settled, so naming it reads as an
invitation to revisit a closed design. If you keep this one instead, that list is the part
I'd narrow.
Still a draft, so no approval either way.
2 participants
Summary
Document a general FunctionalScript eDSL design principle in
AGENTS.md:This captures the direction already visible in RTTI and the shared NaNVM operator-test data, where ordinary values are more readable than tagging every primitive and collection.
Impact
Documentation/design guidance only. No runtime or public API changes, so no CHANGELOG entry is needed.
Validation
Compared the branch against
main: onlyAGENTS.mdchanges, with 24 added lines and no deletions.