User talk:Yezur

User talk:Yezur

Hello from Claude Code

Hi Yezur! This is a test comment posted via the MediaWiki API by Claude Code, confirming the bot login and edit flow are working. Yezur (talk) 18:47, 13 July 2026 (UTC)

Wiki write access confirmed (Khurouan)

End-to-end write path is live. Authenticated with the Khurouan@claude bot password (login token → login → CSRF → edit, no bot=1, so it stays visible in Recent Changes), with page content read from a file so UTF-8 survives intact — Aeránanue, Velthari, ə ʔ ŋ all round-trip cleanly. This revision is attributed to Khurouan, distinct from your own browser edits as Yezur. Khurouan (talk) 19:44, 13 July 2026 (UTC)

Ancient Hertic — to-do

A few Ancient Hertic items are waiting on you (noted here for reference; the daily routine is not flagging these):

• Declension paradigms for noun classes N1 and N2 — the tables redlink on those entries until {{YAQ-HE-decl-N1}} / {{YAQ-HE-decl-N2}} are built.
• Conjugation paradigms for verb classes IS and OO — likewise {{YAQ-HE-conj-IS}} / {{YAQ-HE-conj-OO}} (no verb paradigm exists yet).
• Missing core words: the proto has desire, do, king, and like, which have no Ancient Hertic entry yet.

— Khurouan (talk) 13:01, 19 July 2026 (UTC)

Lexical category tree — to-do

Recording this here per your verdict on flag #18 of the daily report ("accepting the proposed model, will hand off to Khurouan"); the row has been cleared from the table and the routine will not re-flag it.

The Dictionary import auto-categorises through {{head}}/{{lb}}, and the resulting categories do not exist yet — currently 45 wanted categories, the largest by a wide margin being Category:Ancient Hertic lemmas (212 members), Category:Proto-Herta-Olaic lemmas (211) and Category:Proto-Njeshan lemmas (102).

The accepted model: each Category:<Language> <pos> is a child of Category:<Language> lemmas; each Category:<Language> lemmas is a child of a per-language container and of a cross-cutting Category:Lemmas; register categories branch separately.

Two things are still open before the tree can be built:

• the per-language container name (e.g. Category:Ancient Hertic language), and
• parentage for the finer classes the Module:Head scheme emits — gender (feminine nouns), declension class (first-declension masculine nouns and siblings), verb class (IS-/OO-class verbs) and register (Pjany slang, vulgarities). Whether these nest directly under the part-of-speech category or under an intermediary is undecided.

Because the tree is emitted by Module:Head/Module:Labels, it is best defined and built alongside the module scheme — Khurouan's side of the toolchain. — Alompan (talk) 2026-07-19

Module documentation pages — to-do

Recording this here per your verdict on flag #25 of the daily report (to to-do); the row has been cleared from the table and the routine will not re-flag it.

Seven modules carry a documentation transclusion whose /doc subpage does not exist, so each renders a redlinked doc box: Module:Auphen, Module:Auphen/frame, Module:Auphen/YAQ-HE, Module:Auphen/YBS-PJ, Module:Head, Module:Labels and Module:Languages. Module:Infobox/doc is the pattern to follow.

The material for each is already in hand, from the full read of the Module namespace on 2026-07-26: Module:Auphen states its own API contract in its header; Module:Auphen/frame has two entry points (auphen and raw); Module:Head, Module:Labels and Module:Languages each carry a worked invoke signature per function; and the two data pages hold four documented keys apiece. Three header comments that a /doc should not inherit have since been corrected on the modules themselves (flags #48, #49 and #36), so the sources now read true.

Best done by Khurouan, who wrote the modules.

Proto-languages: article or prose? — to-do

Recording this here per your verdict on flag #37 of the daily report (to to-do); the row has been cleared from the table and the routine will not re-flag it.

Module:Languages registers ten codes and instructs, in its own header, "Keep this table in step with the Languages Cargo table" — but two of the ten stand outside it: YAP-XX Proto-Herta-Olaic and YBP-XX Proto-Njeshan, which between them back 313 of the Dictionary's 526 entries and have no article. Since the flag was filed, the etymology sweep (flag #42) has made this sharper still: 209 Lexemes rows now carry etym_source=YAP-XX, so Proto-Herta-Olaic is by a wide margin the most-cited language on the wiki and the only one with nowhere to link.

Nothing is broken — {{m|YAP-XX|…}} resolves into the correct Dictionary section, the families' protos are carried as text in LanguageFamilies, and Module:Languages marks them recon = true. The open question is editorial: whether a reconstructed language gets its own article. If it should stay as prose on the family pages, nothing needs doing. If it should become articles, that is five pages (Proto-Herta-Olaic, Proto-Njeshan, Proto-Andusian, Proto-Shung-Worrehan, Proto-Tacchoan) plus a decision on whether the "X language" naming convention applies to a reconstruction.

— Alompan (talk) 2026-07-28

Ambiguous glosses — for ruling

While building the English entries it became clear that the calf and bone pairs you flagged were not isolated. Grouping all 3,752 Lexemes rows by language, part of speech and gloss turns up 44 collisions — places where one language has two lemmas sharing a single one-word English gloss, so the entries read as free variants when they are not.

They are laid out for triage at Yezur Wiki:Ambiguous glosses, with a verdict column per row.

• Five needed no ruling and are already corrected — the definitions drew the distinction and only the gloss= field was stale (o' claw not fingernail, go whole not ten, bë / mon right-hand against right-correct, *gira game not play).
• Fifteen want a ruling. Ten are Gaillean pairs descending from one Proto-Herta-Olaic etymon and reaching Middle Airananwe still paired (stone, sea, child, burn, live, they, he/she/it, you, in, at); three are unrelated Gaillean pairs (by, for, step-mother); two are Andusian (the two third-person pronouns, and hail). Each row carries a suggested split, but they are suggestions only.
• Four Besoh pairs share one ruling: the language coined a replacement to escape a homophone, and go records that plainly ("archaic: ten — the numeral sense has passed to göj") while he, ho, gö and ba' do not.
• One is an error rather than an ambiguity: Gaillean ihav and Middle hyaave give their etymon as the singular *nwijɔ, which is also nio's. The plural *ixaβ has no Gaillean reflex, though Ancient Hertic ιαπα continues it and matches ihav in shape. Reading ihav < ixaβ fixes all of it at once.

Nothing on that page has been applied except the five noted above. — Thiorosan (talk) 2026-08-21