Rebuild the NVDA sample holon on xbrlkit 0.7.2 - #77
Merged
Conversation
The sample the viewer opens on load — NVIDIA's FY2026 10-K — was built on 2026-07-30, before the holon carried what it carries now. Rebuilt from the same filing on xbrlkit 0.7.2. Same report: 1,425 elements, 1,219 facts, 2,392 associations, 92 dimensions, 89 information blocks, 15 periods, 6 units — the parse is unchanged, only what the serialization says about it. What the file gains: - **344 elements that did not compact at all.** NVIDIA's own concepts were written as `https://robosystems.ai/concept/nvda:A2026NotesMember`, because the emitter had no binding for a filer's taxonomy. They are now `nvda:A2026NotesMember`, bound to `http://www.nvidia.com/20260125#`. - **The real taxonomy namespaces.** `us-gaap` resolves to `http://fasb.org/us-gaap/2025#` rather than a year-less stem no filing uses. - **Every label role.** The elements carried `prefLabel` alone, so the viewer rendered standard labels rather than the ones NVIDIA chose — the 2026 notes read "2026 Notes [Member]" where the filer's terse label says "3.20% Notes Due 2026". `terseLabel`, `totalLabel`, `documentation` and the rest are now on the elements, with their language. - `dataType`, `baseType`, `nillable`, `language`, `isNil`, `periodEndDate`. It is ~11% bigger (4.9 → 5.5 MiB): the compaction saves bytes and the label palette costs more than it saves. Nothing reads differently — the viewer names concepts by `rs:internalId`, which is unchanged — and all 51 tests pass, including the two that read this file.
jfrench9
added a commit
that referenced
this pull request
Sep 8, 2026
The sample refreshed in #77 was built on xbrlkit 0.7.2, which bound `label` to `rs:label` — a duplicate key in its canonical context silently shadowing the `rdfs:label` that was already there. It went live with the 06:13Z deploy, so the file the viewer opens to show what a holon *is* has been disagreeing with every platform-written holon about what one of its terms means. Rebuilt on xbrlkit main with that fixed (xbrlkit #37): `label` is `rdfs:label` again and the standard-role label is `standardLabel`, XBRL's own name for it. Only the term changes — same 1,425 elements, 1,219 facts, 382 networks. The viewer itself was never affected: it names concepts by `rs:internalId`, so it rendered correctly throughout. What was wrong is what a third party reading the file as JSON-LD would resolve, which is precisely what a reference sample invites someone to do. Built from the merged fix rather than a published version — a 0.7.3 would make this reproducible from a pin.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The sample the viewer opens on load — NVIDIA's FY2026 10-K — was built on 2026-07-30, before the holon carried what it carries now. Rebuilt from the same filing on xbrlkit 0.7.2.
Same report, so this is not a content change: 1,425 elements, 1,219 facts, 2,392 associations, 92 dimensions, 89 information blocks, 15 periods, 6 units — identical on both sides. Only what the serialization says about it changed.
What the file gains
https://robosystems.ai/concept/nvda:A2026NotesMember, because the emitter had no binding for a filer's taxonomy. They're nownvda:A2026NotesMember, bound tohttp://www.nvidia.com/20260125#.us-gaap→http://fasb.org/us-gaap/2025#, not a year-less stem no filing uses.prefLabelalone, so the viewer rendered standard labels rather than the ones NVIDIA chose. The 2026 notes read "2026 Notes [Member]" where the filer's terse label says "3.20% Notes Due 2026".terseLabel,totalLabel,documentationand the rest are now present, with their language.dataType,baseType,nillable,language,isNil,periodEndDate.It's ~11% bigger (4.9 → 5.5 MiB): the compaction saves bytes, the label palette costs more than it saves.
Compatibility
Nothing reads differently — the viewer names concepts by
rs:internalId(asdescribeReport.tssays in its own header), and that is unchanged. All 51 tests pass, includingsmoke.test.tsandholon-chat.test.ts, which both read this file directly. Prettier clean.🤖 Generated with Claude Code
https://claude.ai/code/session_01EiGTpXmxwZjxzaRoVs4VxB