--- name: ARC rework landscape (reawakening 2026-04-16) description: ARC work surfaced after local dev environment restored. Captures the landscape as articulated by the steward — content rework, versioning question, cleanup scope. Not a project plan; a list of what is seen and named, to be taken up when concrete. type: project originSessionId: 19aac84d-35aa-466d-bdc8-6d7c07cf449a permalink: claude-memory/project-arc-rework --- # ARC rework landscape — from 2026-04-16 reawakening After local dev was restored on 2026-04-16, the steward returned to ARC with fresh eyes for the first time in months and named the following scope. Preserved in their voice; not yet acted on. ## DECISION 2026-04-19: ARC stays unified — about page reflows (hic first, *Adsum* later) The decision evolved in three moves within a single day: 1. **Separation rejected.** Steward considered ARC ↔ davidglidden.eu split earlier in the week; reconsidered and rejected. Cross-discipline integration **is** what the steward is; a separated personal site would perform a partition that doesn't exist in the work. 2. **Two-peer-compass split briefly adopted, then rejected.** Considered splitting the current "About" Reading Compass point into two peer compass points (Colophon for artifact + *Adsum* for person). Rejected on the same grounds as separation: any tidy artifact-vs-person partition re-performs the separation the steward rejected, just at a smaller scale. 3. **Arrived-at architecture: single about page, reflowed.** The **hic** register (about-the-artifact: apparatus, formation, design, tools, typography, navigation philosophy) comes first; ***Adsum*** register (about-the-person: biographical, availability, concerts, mentorship, pedagogical research) comes later in the same page. One page, multidisciplinary by embrace rather than by partition. **Steward's framing:** *"embracing the multidisciplinary aspects of my work rather than trying to separate them tidily."* ### Why reflow, not split Every split re-performs the partition the steward rejected. The ARC-unified decision said cross-discipline integration is what the steward is. The two-peer-compass shape is the same move at a smaller scale — artifact-self and person-self as tidy adjacents. Reflow lets them coexist in one page, in an order (hic → *Adsum*) that honors ARC's compose-the-work-first grammar without announcing "here is apparatus, here is person" as a structural claim. ### *Adsum* name carried forward as register, not compass point Though no longer a peer compass point, the ***Adsum*** naming is retained as the register-name for the person-facing section of the reflowed about page. The rationale still holds. *Adsum* is Latin for *"I am here / I am present"* — the Roman answering-word, the response when one's name was called at a roll. It marks the projected image as a *response* to being called, not a self-presentation. Resonances: - Pairs with *hic jacet* in the existing apparatus grammar (work lies here; person is present here) - Resonates with L1's I15 (the Accusative Default — answerable, not sovereign) - Avoids "About me" / "Bio" / "Vita" (sovereign-self / web-2.0 register) - Avoids *Persona* (Jung-freighted in modern English) and *Vita* (CV-coded) - Acknowledges projection-as-projection rather than performing self-presentation (the Harrison-shaped concern: name should refuse the noun-as-self and reach for a verbal/relational form) Plural alternatives held but not chosen (still useful as fallbacks if *Adsum* doesn't wear well): **Callings**, **Vocations** — Harrison-flavored, plural acknowledges multiplicity (musician + writer + teacher + father) without forcing synthesis; you're *called to* these, you don't claim them. Steward's verbatim adoption: *"I adopt adsum."* ### Reasoning that defeated separation (still operative, generalized to defeat peer-split too) - Cross-discipline integration is the substance of the person (violist + writer + father + teacher); any tidy partition performs a separation that doesn't exist in the work - **Gwern.net as steward-referenced precedent.** Pluralism handled by *substance* (what's written), not navigation gymnastics. Biographical kept secondary (footer `/me`, not primary nav). Identity stated by writings, not credentials: *"This is the website of Gwern Branwen. I write about AI, psychology, & statistics."* The "best known for" framing carries multi-discipline identity without ranking. ARC parallel: *"Animal Rationis Capax. The writing of David Glidden."* - What ARC takes from gwern: declarative identity by writings; biographical kept secondary; pluralism native to the substance. What ARC does not take: link-dense ToC homepage, multi-column layout, topic-prefix naming ("AI:", "QS:") — violates ARC's editorial breath. - One difference from gwern: gwern has no concerts to schedule. Steward does. Inquirers (orchestras, students, mentees) need a findable surface. In the reflow shape this is served by *Adsum* section presence on the about page, reachable via the About compass point — not by elevation to a peer compass. - Less infrastructure debt: one site, one build, one deploy, one about page - The contemplative register isn't violated — the about page already lives in a different register from the essays ### URLs / DNS ARC moves to **animalrationiscapax.org** via Cloudflare Pages. **davidglidden.eu** becomes a redirect pointing at ARC's new home — not a separate site. The personal domain becomes a pointer, consolidating the public surface under animalrationiscapax.org. **Migration state (2026-04-19 evening, terminal-crash recovery):** - Porkbun (registrar): nameservers updated to Cloudflare (`priscilla.ns.cloudflare.com`, `tom.ns.cloudflare.com`). whois confirms. Porkbun's NS `REFUSED` when queried directly — they've released the zone. ✓ - Cloudflare: zone status **Active**. priscilla holds the zone authoritatively with CF-signatured SOA. ✓ - Stale apex A records (`44.227.76.166`, `44.227.65.245`, AWS Oregon IPs — origin unclear, probably import artifact) deleted by steward. ✓ - Public resolvers (1.1.1.1 / 8.8.8.8 / 9.9.9.9) still return cached porkbun NS at time of diagnostic. Propagation of the TLD delegation takes up to ~24h due to standard .org NS TTLs. The parking page still appears via this cached path — **this resolves on its own, not a misconfiguration**. Independent verification (from outside steward's network) available at dnschecker.org. - SOA primary reads `curitiba.ns.porkbun.com` (stale imported value, usually overridden by CF as zone fully settles). Not breaking. Re-verify later. **Next steps for the move:** - **Bind `animalrationiscapax.org` as custom domain in the ARC Cloudflare Pages project** — this auto-creates correct apex records (CNAME flattening to `.pages.dev` or proxied A records). The deletion of the old A records + the absence of the Pages binding is why an apex lookup via CF currently returns nothing. - Verify propagation after a few hours: `dig @1.1.1.1 NS animalrationiscapax.org +short` should show CF nameservers. - Configure davidglidden.eu → animalrationiscapax.org redirect at whichever layer is fitting (CF Page Rule if it's also on CF; registrar-level URL forwarding otherwise). - Post-propagation: update ARC's internal canonical URL (Hakyll `siteRoot` in `site.hs`, canonical `` tags, RSS feed URL, sitemap) to `https://animalrationiscapax.org`. Verify `make rebuild` + deploy picks up the new origin; CF Pages deploy target may need reconfiguration for the new custom domain. - Update Acerplex push path / any other CI references to the new URL. *(In progress at time of terminal crash, 2026-04-19 evening.)* ### Resolved by this decision (formerly open work) - `davidglidden-eu` repo scaffold — moot (no separate personal site) - Separate Hakyll build for davidglidden.eu — moot - Cross-link convention design — moot - Colophon as a separate compass point — moot (Colophon-register content flows as "hic first" within the about page; name *Colophon* may still serve as a section heading, open) - Glyph design for an *Adsum* compass point — moot (no new compass point) - Compass optical rebalancing to accommodate a 10th point — moot - About-enfilade split between Colophon and *Adsum* — moot (single page; enfilade nav to reflowed about only) - About-pages biographical content (Desk Interlude with Kai unborn, Lune presence, Manutius paraphrase) — reshapes from "rewrite under *Adsum*" to "compose *Adsum* section within the reflowed about page" ### Open work generated by this decision - **Reflow the existing about page**: arrange hic-register content (apparatus, formation, design, navigation philosophy, typography) first; compose *Adsum*-register content (bio, concerts, mentorship, pedagogical research, coaching availability) as later sections of the same page. Substantial content work; not infrastructure. - **Desk Interlude update**: Kai-as-unborn → Kai-born; Lune now present in the room. Same page, updated text. - **Manutius paraphrase removal** + too-many-quotes prune — same page. - **Bio + Resume treatment**: integrate into the *Adsum* section rather than moving to a separate destination. - **hic / hic jacet philosophical commitment** (from vignette spec) gets embodied in the reflow, not announced as a feature. - **DNS / Cloudflare move** — in progress at crash time: ARC → animalrationiscapax.org (via Cloudflare Pages), davidglidden.eu → redirect. Resume from that state. - **Colophon retained as section heading** (confirmed 2026-04-19) for the hic-register content within the reflowed about page. - **Open design question — subtle *hic* register marker.** Steward's intuition 2026-04-19: if the hic section is simply titled *Colophon* and *Adsum* arrives later without context, the register-shift may read as abrupt. The *hic* register itself may want a quiet identifier — not a headline, something subtler (a small-caps lead-in, an italic phrase, a tiny apparatus mark) that does the register-declaring work without announcing the distinction structurally. The risk to avoid: re-performing the artifact-vs-person split at the micro-scale through heavy signposting (that's what the peer-compass architecture was rejected for). The reflow should feel continuous even while the registers modulate. **Mockup resolution 2026-04-19 evening**: the section names *Colophon* (about-artifact) and *Adsum* (about-person) do the register-marking themselves; no added italic subtitle needed. Names serve, added subtitles start to re-perform. ### Mockup reviewed 2026-04-19 evening (Option A) A full draft of the reflowed about.md was produced for steward review. Decisions locked: - **Adsum opening drop-cap paragraph** ("I answer to my name here—not as self-presentation but as response...") — KEEP. Embodies the register-declaration inside the page rather than announcing it as a feature elsewhere. - **Seasonal Geometries section** — MOVE to `navigation-philosophy.md`. Reading Compass is its own entity; seasonal glyphs belong with the navigation explanation, not with the about-page flow. - **Henry James quote placement** — STAY at close of page (after Contact), per steward's no-strong-preference. Open from mockup review: - **Drop-cap policy question (Warde tension).** Steward on fence. Warde's *Crystal Goblet* principle: type should be transparent; drop caps may read as performative decoration. Vignette system (on essays) adds another threshold; drop cap + vignette + dateline breadcrumb = too many opening signals. About pages won't have vignettes, so the About case is different from the essay case. **Executor lean: option (b)** — keep only the Adsum drop-cap, drop the site-intro drop-cap. Forward rule: *drop caps mark register-declarations, not page-tops.* Emerson block already opens the page; the site-intro drop-cap stacks a redundant threshold. Adsum drop-cap marks a register-shift inside the page, which is semantically load-bearing — passes Warde's "the form does work" test. Not yet decided. - **About-pages classification problem** (steward raised 2026-04-19 evening, **new**). All `content/pages/*.md` currently carry `class: essay`, which pulls essay-apparatus (breadcrumb formatting, vignette slot once live, essay-style datelines, probably reading-time) onto pages that aren't essays. Creates confusion in the breadcrumb system at top of each page. **Proposed solution: introduce `class: about` as first-class content type**, matching the existing ✍︎ About Reading Compass point. Migration: every file in content/pages/ (about, colophon, bio, resume, aldine-xxi, typography, navigation-philosophy, vignette, attention-page, influences-page, tools, patterns-page, now-page, council-page-public if still relevant). Template treatment for `class: about`: no vignette, different breadcrumb (or none — enfilade nav already lives at page bottom), versioning infrastructure retained (ADR-005 is class-agnostic — renders breadcrumb when `revisions:` or `edition_label:` present). Scope is small but nonzero: `site.hs` has class-switch(es) that need an `about` branch; templates may need an `about.html` variant or conditional in `default.html`. **Steward agreement pending on approach + scoping.** Recommendation: scope as its own PR rather than bundling into about-reflow PR — different concerns (content flow vs. infrastructure), cleaner commit, useful independent of whether reflow lands. ### Triggering question carried forward *"How does one mark the hic of one's projected image to others?"* The reflow is the answer at the level of the page: hic precedes *Adsum*; the thinking-here is declared before the presenting-here; the work is foregrounded before the person. Mark the public surface as projection. Don't pretend it's the self. *Adsum* is the answering-word that acknowledges being-called rather than self-presenting. Future work that touches *Adsum*'s shape should be evaluated against this commitment. --- ## Codification surface — what warrants spec-level treatment (2026-04-19) Question surfaced during the about-page reflow session. The `class: essay` fallback for everything in `content/pages/` is a *symptom*, not the problem: **ARC's content typology itself is underspecified**. The about-page case is one symptom; the "genre riddle" from the mobile review (meditation / observation / fragment overlap) is another. Steward reframed: what ELSE in ARC would benefit from codification? ### Test for when codification is warranted Codify when more than one holds: - **Recurs** — the same judgment is made many times - **Has downstream consumers** — templates, skills, agents that apply the rule - **Embodies a commitment that should resist drift** — philosophical load-bearing - **Discovered, not invented** — the rule emerged from practice and is stable Don't codify what should stay live judgment (unresolved aesthetics, in-flux questions, fashion). ### Strong candidates 1. **Content typology / genre system** (THE recommendation). Defines each type (essay, observation, meditation, fragment, glimpse, photo-essay, gloss, about, chamber), resolves overlaps, defines per-type apparatus (vignette yes/no, breadcrumb style, dateline shape), binds to `class:` frontmatter + Reading Compass glyph. Downstream consumers: `site.hs` routing, semantic-analysis skill, vignette system, style guide. Spec-level precision warranted (peer of AldineXXI codex and vignette specification). The `class: about` question is one entry in this spec, not a separate problem. 2. **Apparatus grammar** (hic / hic jacet, dateline, location, edition labels, revision breadcrumb). Partially in ADR-005, scattered in vignette-specification, implicit in colophon prose. Worth consolidating as its own Codex page — "Grammar of Place" or similar. Builds on 2026-04-13 hic philosophical commitment. 3. **Pattern 3 typography decision** (2026-04-18). Decided but living in a memory file. Move to style guide when that session happens, or brief ADR sooner. The decision is stable; memory-only is undercodified for a load-bearing commitment. 4. **Reading Compass rules.** Currently declarative on About; spec form would define *what qualifies* (when ● vs ○ vs ◈ — the "quality of attention" criteria), not just what the glyphs label. Feeds directly into typology (§1 catalogue). ### Medium candidates - **Style-guide prose positions** (UNESCO / MLA / Chicago declared positions) — named in mobile review list as needing redevelopment. Downstream of typology; should follow it. - **Navigation hierarchy rules** (compass vs enfilade vs breadcrumb vs post-nav vs version-stamp) — implicit but operative. Partially covered in navigation-philosophy.md as prose; spec form would codify the *rules* rather than the philosophy. ### Later candidates - **Semantic-analysis skill spec** — premature until content typology is locked (skill consumes typology). ### Recommendation and sequencing **Content typology spec first.** Foundation; everything else is downstream. Reasons: - Resolves `class: about` cleanly (the class becomes one field in the typology) - Unblocks the genre-riddle cleanup in the mobile review list - Feeds the style-guide redevelopment (item in mobile review list) - Prerequisite for the semantic-analysis skill - Tractable in a single focused session **Home for the spec when authored:** `docs/AldineXXI-Codex/content-typology-specification.md` (peer of `aldine-xxi-specification.md` and `vignette-specification.md`). **Status 2026-04-19 evening:** **Draft saved** to `~/_Dev/animal-davidglidden-eu/docs/AldineXXI-Codex/content-typology-specification.md` (untracked; steward commits when ready). Contains §0 organizing principle (multi-axis typology) · §1 Type catalogue (structure + initial table + locked entries for Photo Essay, Glimpse, Fragment/Observation/Meditation drafts) · §2 Apparatus binding (structure only; matrix deferred) · §3 Qualifying marks · §4 Overlap and resolution · §5 Code binding contract · §6 Change policy · §7 Open questions · Changelog · Proposed authoring sequence. Session-locked decisions: multi-axis principle, Photo Essay definition (Berger *Ways of Seeing*), Glimpse:Photo-Essay :: Fragment:Essay pairing, Meditation promoted to 10th type, drop-cap rule (register-shifts only), no-essay-fallback rule. Open at pause: Meditation glyph (steward flagged 2026-04-19); remaining §1 entries for extant types; §2 apparatus matrix content; Poem / Council-Deliberation / Archive-class / Gloss-boundary questions. **In-session decisions locked for the spec (as of 2026-04-19 evening):** - **Photo Essay definition (§1 + §4), sharpened 2026-04-19 by steward:** **An essay whose medium is the image sequence.** The sequence itself constructs the argument, rhythm, turn, and return — the essayistic work is done visually. No interleaved prose between images. Text, where present, is apparatus only (title, introduction, postface, per-image captions, dateline) — does not carry argumentative weight. **Canonical external exemplar: John Berger, *Ways of Seeing*** (the wordless image sequences constructing visual-critical arguments). The boundary with Essay: both are essayistic; they differ in medium (prose vs. image sequence). An essay may embed images as illustration; a photo essay may not embed prose as commentary. - **Substance-primacy axis, refined:** types divide by *which medium carries the argument*. Visual-primacy types (glimpse, photo-essay) construct their substance in image(s); text is apparatus. Prose-primacy types (essay, observation, meditation, fragment, gloss, about) construct their substance in prose; images support. The axis isn't "where text lives" — it's "what medium carries the argument." The steward's Ways-of-Seeing anchor clarified: photo-essay is essayistic all the way down; just in a different medium. Proposed as spec introductory frame. --- > **Sections below this line predate the 2026-04-19 decision.** The "Architectural: decouple ARC from personal domain" section and the "TODO: scaffold davidglidden-eu repo" subsection are SUPERSEDED — kept for the reasoning record. The biographical content surfaced in "About-pages constellation — needs rework" reshapes into *Adsum* authoring rather than ARC cleanup. --- ## Essay versioning — the philosophical question (TODO: explore together) > "The after-the-reply sequence, of which the first two in their original forms are published, but they've been updated since — I need to think about how to deal with essay updates: how to remain true to the text, to have the old versions still available, to let them be superseded?" Explicitly marked to-do. To be explored together, not resolved quickly. The answer shapes what an ARC essay *is*: - Montaigne's additive layers (each revision adds; original preserved within) - Talmudic-style commentary strata - "Revised edition" as its own artifact (original stays as it was; revisit becomes its own piece) - Supersession with git-preserved history (only latest is presented; original exists in archive) - Revision notes / scar tissue inline Downstream implications: whether "sequences" are a navigation layer or a deeper commitment to plural time. Per this listing order. ## After-the-reply sequence - First two essays: published in original form - Both updated since (see versioning question above) - Rest of the sequence: needs publication - Sequence needs UX: either a new category, a taxonomy tag, a frontmatter field, or a Reading Compass glyph. Currently not first-class in ARC's 8-point compass. ## Architectural: decouple ARC from personal domain (2026-04-16) > "The time has come to decouple ARC from my personal domain and to build a light site just for biographical and possibly very specific musical publishing: the pedagogical research observations, the mentorship program, things like that. ARC is a different place. Some work may be cross-published, but ARC needs to be its own entity." Named as a decision-in-progress. Key facts: - Steward owns `animalrationiscapax.org` — ARC's future home. - `davidglidden.eu` becomes a light biographical + musical/pedagogical site (pedagogy observations, mentorship program, etc.). - ARC is "its own entity" — not a subspace of David Glidden the person. - Cross-publishing possible for some pieces. Implications surfaced (not resolved): - **Resolves part of the about-pages rework.** Biographical content (Desk Interlude with Kai/Lune, etc.) likely belongs on `davidglidden.eu`, not ARC. - **Compounds the versioning question.** If a piece lives on ARC and is also cross-published to `davidglidden.eu`, what's the relationship between the two instances? Canonical URL? Does the cross-pub freeze at its pub date or follow ARC's evolution? Becomes a second axis of the versioning discussion. - **Technical ordering (not blocking today):** DNS + CloudFlare move for ARC → animalrationiscapax.org; new Hakyll site for davidglidden.eu (lighter template, probably a fork); git repo rename on Gitea (cosmetic); content triage (what stays on ARC, what moves, what cross-pubs). ### TODO: scaffold `davidglidden-eu` repo (personal site) **When:** once the steward is ready to begin the split, not before. **Shape:** NOT a GitHub-style fork (which preserves ARC's commit history as lineage). A derivation — new repo with fresh git history, seeded by copying in selected files from ARC, then `git init` + first commit: "Initial personal site scaffold (derived from ARC)." **Proposed location:** `git.skemantix.com/David/davidglidden-eu` (on the same Gitea as ARC). New CloudFlare Pages project. **Inherit from ARC:** - Stack / GHC toolchain config (`stack.yaml`, `package.yaml`, `*.cabal`) - Makefile (with deploy targets adapted to the new CF Pages project) - `scripts/deploy-cloudflare.sh` (with a different `--project-name`) - `scripts/deploy-mirror.sh` (with a different mirror repo target, if a mirror is wanted at all) - SCSS AldineXXI foundation (typography, palette, spacing scales) — imported without chamber/triptych/glyph-specific partials - Fonts (EB Garamond, IBM Plex) and `@font-face` declarations - `.gitignore`, pre-commit hooks setup (via dotfiles `core.hooksPath`) - CLAUDE.md pattern (adapted to the simpler site's scope) **Strip from ARC:** - Reading Compass 8-point navigation (replaced with a simpler nav: bio / works / teaching / contact / writing → link out to ARC) - Chamber (canon, voices, library) — has no place on the personal site - Glyph system (GenomeConceptExtractor, archetype selection, semantic visual translator) - Triptych renderers + all associated JS - UCE/CIE infrastructure (already removed from ARC itself in `b1f28d3`, so naturally absent) - Observer-problem, OP-series content - Most of `site.hs` — probably ~70% drops out; what remains is: posts, pages, about, RSS, simple index, CSS compile - `content/chamber/`, `content/fragments/`, `content/glimpses/` (unless steward wants glimpses on the personal site — decision) **Open questions to resolve at fork time:** - Does the personal site need a GitHub Pages mirror at all, or is CloudFlare Pages alone sufficient? - What's the naming: `davidglidden-eu` (current ARC pattern, matches domain) or something else? - Which `content/` types carry over — just `pages/`, or also `posts/` (as a writing/journal surface)? - Cross-publishing mechanism: manual copy, or a content sync (Git submodule / symlink / scripted) — probably manual initially, can evolve. **Sequencing note:** the fork should happen *after* the steward has decided the versioning question (at least provisionally), because cross-pub workflow depends on it. Not yet a plan; a direction named. ### Update 2026-04-18 evening (steward, during mobile-review enfilade discussion) > *"this is a bigger project (which should be noted in memory somewhere) which entail[s] separating ARC from my personal site. Once the separation is made, my personal site will have my biographical info, concerts, mentorship, perhaps technical musical essays (but that depends), perhaps some cross posted material — but things that pertain directly to myself. ARC needs to be independent now. I would link to it from my personal site by mentioning that I am the creator and author of ARC."* **New/sharpened:** - Explicit inclusion of **concerts** and **mentorship** in personal-site scope - **Technical musical essays** "depends" — open question - **Cross-linking pattern:** personal site mentions David as creator/author of ARC (no assumed reciprocal link from ARC to personal site — keeps ARC's reticence register where even children are "my son" rather than named) - Stronger framing: *"ARC needs to be independent now"* — decoupling is active intent, not eventual **Immediate consequence decided 2026-04-18:** Bio and Resume out of the about-enfilade. Treat ARC as the independent entity it's becoming, even before davidglidden.eu exists. About.md's Contact section will eventually redirect bio/resume to davidglidden.eu; transition state TBD. ## About-pages constellation — needs rework Specific items named: - **Desk Interlude**: Kai is still "unborn" in the text. Needs updating (Kai is born). - **Lune's presence**: she has now been in the room. Her presence in the about-pages changes further. - **Too many quotes** throughout. - **Manutius paraphrase**: has to go. - **Hic / hic jacet philosophical commitment** (from vignette spec, 2026-04-13) needs to be integrated into the about pages — not announced as a feature, but embodied. ARC's grammar of place (threshold objects opening and closing each piece as *this thinking happened here, in this form, at this time*) is constitutive, not decorative. The about constellation should reflect this. - "etc etc etc" — more to come as revision proceeds. ## Other pieces queued - A poem to publish. - Lots of observations. - Lots of glimpses. - *bref.* ## TODO: Semantic analysis agent/skill — AldineXXI application + vignette seed generation Named 2026-04-16. Not yet designed. **Purpose:** take a publish-ready piece → (1) analyze semantically → (2) apply/recommend AldineXXI typographical conventions → (3) generate the seed data for the vignette system. **State of prerequisites:** - AldineXXI v2.0 spec: exists (`docs/AldineXXI-Codex/AldineXXI-specification.md`) - Vignette spec: exists (`docs/AldineXXI-Codex/vignette-specification.md`), jurist-approved - Vignette renderer: **not yet built** — Phase 3 pending (waiting on Gitea restart, fresh session, Hakyll build — first two now resolved today) - Gap map: `docs/AldineXXI-Codex/gap-map-2026-04-13.md` **Key design questions (unresolved):** - **Agent vs skill vs both?** Claude Code skill (interactive, authored run before publish) vs autonomous agent (runs on build/CI) vs hybrid (skill for authoring, agent for validation). The steward phrased it as "agent or skill" — decision open. - **Where does it live?** A Claude Code skill on the steward's machine? A Haskell component in ARC's site.hs build? A separate tool invoked from Makefile? Three different postures with different lifecycle implications. - **When does it run?** Pre-publish interactive (author-controlled), on Hakyll build (automatic, no author decision), or both (skill produces annotations that build consumes)? - **Relationship to existing glyph system** (`assets/js/glyphs/`, preserved in `b1f28d3`)? Parallel, integrated, or eventually replacing? The glyph system is client-side visual output; this new agent is build-time semantic pre-processing. Potentially complementary. - **Potential Chamber/BMF integration** (later): Chamber's voice recognition + BMF recall could inform semantic analysis (e.g. "this passage echoes Levinas with confidence 0.8 against chamber-library chunks"). Premature coupling risk — start standalone, let integration emerge from practice. **Inputs (provisional):** - Essay content (markdown + frontmatter) - AldineXXI v2.0 spec (as knowledge base / rules) - Vignette spec (as schema for seed output) **Outputs (provisional):** - Typographical annotations or transformations (small caps, drop caps, optical spacing, ligature hints, voice attribution, cite styling, em-dash discipline) - Vignette seed: structured data the vignette renderer will consume to produce the per-piece visual element **Dependency note:** building this before the vignette renderer exists means we design the seed format somewhat speculatively. Probably wise to build vignette renderer first (or at least stub it), then build the agent to produce seeds the renderer can consume. Or design them together. **Architectural observation (2026-04-18):** the skill has three layers with different semantic requirements — and the third overlaps structurally with the vignette engine's analysis component. 1. *Fully mechanical* — drop-cap on first body letter, trailing double-spaces, ornament divs at section boundaries, hard-break discipline. CSS rules or markup discipline. No analysis needed. 2. *Lightly linguistic* — small-caps on proper names (standard NER). Shallow, bounded, doesn't need the vignette's depth of reading. 3. *Expressive* — paragraph-class decisions (`pull-quote` / `epigraph` / `breathing-room`), ornament *variation* between sections (centripetal essay gets one style, distributed essay another), drop-cap glyph choice, pacing decisions honoring the essay's shape. This is where "organizing logic, philosophical character, where the text concentrates vs opens" matters — exactly the reading the vignette's Phase 2 generation script (semantic analysis → genome JSON per `vignette-specification.md §Part II.Component 1`) needs. Design implication: the expressive layer of the typography skill and the vignette engine's semantic analysis component should share *one* semantic analyzer. Built once, consumed twice. Whether they share output schema or just the analysis pipeline is an implementation decision for the design session; the constraint is they should not independently invent their own semantic readers. This supports the prime directive (*do things once, correctly*) and mirrors the architectural pattern the steward already uses in L2 (one substrate, multiple governance consumers). ## TODO: Style guide redevelopment (noted 2026-04-18) Existing style guide at `docs/style-guide.md` (527 lines) is substantial and comprehensive on ARC's *internal* vocabulary — semantic post types, ornament classes, Chamber citation notation (°, ~, †, §, ∞, ※), Reading Compass, forbidden patterns, small-caps conventions. Stale on multiple fronts: - Pre-ADR-005: no versioning regime, no hic/hic jacet philosophical commitment, no retractation-piece notation - Pre-AldineXXI-codex cross-reference: doesn't point to `docs/AldineXXI-Codex/` specifications - Pre-vignette system: still references triptych/glyph system as if operative - Pre-sequence work: no treatment of sequences as first-class More importantly, **does not declare explicit positions on the classic tensions** between UNESCO / MLA / Chicago-Turabian that the steward named. These are the typical style-guide questions that need positions: - Citation format (notes-bibliography vs author-date vs parenthetical vs in-prose) - Quotation conventions (American "" vs British '' for outer quotes; single/double nesting) - Title formatting (italics vs quotes for short/long works; translation conventions) - Date format (`2025-06-17` ISO vs `June 17, 2025` MLA vs day-month-year UK) - Gender/pronoun conventions (generic they vs he-or-she; inclusive pronouns in quoted historical material) - Oxford comma (declared yes/no) - Em-dash vs en-dash vs hyphen (spacing, semantic use) - Ellipsis (three periods vs Unicode ellipsis; bracketed for omissions) - Foreign language treatment (italics always, only on first use, or lang-attribute-only) - Numbers (spelled-out threshold; oldstyle vs lining) - Proper nouns / corporate names / honorifics ARC's actual practice is a principled blend — the guide should make the positions explicit rather than implicit. **Recommended approach (practice-first, thin, explicit):** 1. Audit a handful of recent essays against each classic question — record actual practice 2. Identify tensions between UNESCO / MLA / Chicago positions where ARC's practice diverges from any single guide 3. Declare explicit positions with brief reasoning (*why* ARC takes this position, not just *that* it does) 4. Keep the rich ARC-specific content (semantic types, ornaments) largely intact — rework is *additive*, not rewriting 5. Cross-reference AldineXXI codex (typography) and ADR-005 (versioning) as separate canonical sources; the style guide is about *prose* conventions, not typography or infrastructure **Dependency to surface explicitly:** The style guide is a **prerequisite for the projected ARC agents** — particularly the semantic analysis / typography application skill noted above. Layer 3 of that skill (expressive decisions: which paragraphs warrant pull-quote/epigraph/breathing-room classes, ornament variation, drop-cap glyph choice) requires the style guide's declared positions as its rule set. The skill *enforces* the style guide; the style guide *defines* what the skill enforces. Without the guide's declarative form, the skill has no stable ground to consult. Sequencing note: the style guide can be developed before the typography application skill is built — in fact, must be, for the skill to have rules to apply. It can also proceed in parallel with ongoing publication work; it doesn't block Essay I & II publication (which uses mechanical + lightly-linguistic typography only, per the three-layer decomposition). Not this morning's work. Dedicated session when the steward is ready. ### Typography register — DECIDED 2026-04-18: Pattern 3 (role-assigned families) Surfaced when the steward noticed a duplicate dateline on the live v2 Ethics essay page, which led to a broader conversation about ARC's typographic system. A decision was reached and is recorded here so the style-guide session has a starting position rather than a fresh question. **Decision (authored 2026-04-18):** ARC adopts a **role-assigned typographic system**: - **Serif — EB Garamond** — body text + headings (content register). All editorial/literary content authored by the writer. - **Sans — IBM Plex Sans** — navigation / wayfinding (between-piece). The Reading Compass, enfilade navigations, between-page orientation. - **Mono — IBM Plex Mono** — apparatus (metadata register). Dateline (hic jacet), technical metadata, any "about-the-artifact" labeling. This is **Pattern 3** in the typology surfaced during the 2026-04-18 conversation — distinct from Pattern 1 (classical humanist unification; all-serif with careful variation) and Pattern 2 (magazine/editorial; serif body + sans headings and nav). **Primary justification — accessibility.** Family-change differentiation is more robust under visual impairment than within-family differentiation. Italic, small caps, weight, and size variations rely on fine details that degrade first when acuity drops; family changes (sans vs serif vs mono) are coarser signals (stroke width, x-height, presence/absence of serifs, proportion) that survive visual degradation. The steward is myopic and now uses progressives; ARC should be legible to readers who share that constraint, and the role-assigned system serves them through the coarse signal. **Secondary justifications.** - Architectural clarity: each register has one job; the reader learns to "read" the type itself as meaning - Screen-primary context: on the web the reader cannot control size/proximity the way a print reader can, so the type system has to do more signaling work than in print - Minimal drift from current state: ARC is already ~90% Pattern 3 in practice; no large migration required **Current font stack — retained for now.** EB Garamond (serif) + IBM Plex Sans + IBM Plex Mono. The steward named that they need time to consider font changes; the *system* is decided (Pattern 3), the *specific families* are refinable without changing the system. **Cleanup implied by Pattern 3 adoption (for a future session):** 1. Delete the dead `heading-base` mixin in `scss/AldineXXI/_mixins.scss` — it sets `$sans-family` for headings, which contradicts Pattern 3 (headings are content, should stay serif) and is already overridden by `_base.scss:39`. The mixin is vestigial. 2. Audit remaining nav surfaces (site header/footer, Reading Compass, tag pills, meta labels, breadcrumbs, form UI): for each, decide whether it is *wayfinding* (→ sans), *apparatus* (→ mono), or *editorial content* (→ serif). Declare each in the style guide. 3. Archive the `scss/AldineXXI/archive-v1/` backup files if no longer needed as reference. **Counter-arguments held but not overriding (for future reconsideration):** 1. **Voice tension.** EB Garamond is 16th-century humanist; IBM Plex is 2010s industrial/corporate (designed for IBM's brand). The three families don't equally *belong* to ARC's contemplative register. If this tension becomes aesthetically costly over time, reconsider the specific families (see *Font swap candidates* below) without changing Pattern 3. 2. **Architecture vs. continuous flow.** Pattern 3 announces "here is content, here is nav, here is metadata" via family shift. A strong typographer would argue Pattern 1's unified voice better serves contemplative reading — the reader stays in one typographic world rather than experiencing register-breaks. If future evidence suggests Pattern 3's breaks are fragmenting rather than orienting, Pattern 1 is the fallback. 3. **Designing-for-self bias.** The accessibility benefit is asymmetric — strongest for low-vision readers, neutral-to-mildly-fragmenting for others. Worth naming so it doesn't get rewritten as "Pattern 3 is universally better" in future reasoning. **Font swap candidates considered 2026-04-18 (not adopted, held for future consideration):** Listed roughly from lowest-regret to highest-commitment: - **Mono: iA Writer Mono / Duo (open source).** Swap IBM Plex Mono → iA Writer Mono. Warmer, more humanist than Plex Mono; designed explicitly for long-form writing; the steward authors in iA Writer so this creates through-line continuity from composition to publication. Small change, clear voice-yield. *Hedge:* verify specific repo + naming (Mono / Duo / Quattro) before committing. - **Sans: Alegreya Sans or Ysabeau (open source).** Warmer humanist sans alternatives to IBM Plex Sans. Alegreya Sans (Juan Pablo del Peral) is designed as companion to Alegreya Serif — different genealogy from Garamond but more literary in register than Plex. Ysabeau (Christian Thalmann / Christophe Caignaert) is newer, designed neutral-humanist. Either is a softer pairing with Garamond than Plex. - **Adobe CC path (if the steward has Creative Cloud).** Garamond Premier Pro (Slimbach) + Cronos Pro (Slimbach, humanist sans) — same designer, voice-aligned pair. Premier is the finest Garamond available (optical sizes, full character set). No mono in the Cronos family; still pair with iA Writer Mono. Free under existing CC subscription. - **Superfamily — Fedra (Typotheque, Peter Biľak).** Fedra Serif + Fedra Sans + Fedra Mono, all designed together. The rare case of all three registers as a unified system. Contemporary humanist, editorial register, designer-run foundry with fair licensing. Paid, perpetual, roughly $300–600 for a full weight set (verify current pricing). **The cleanest voice-coherent answer if budget allows.** - **Superfamily — FF Scala + Scala Sans + separate mono (FontFont / Martin Majoor).** Majoor's 1990s classic, among the earliest deliberately-paired serif-sans superfamilies and still among the best. No dedicated mono; pair with iA Writer Mono or similar. Most classical-humanist choice. Paid through Monotype. - **Superfamily — Adelle + Adelle Sans + Adelle Mono (TypeTogether).** True superfamily spanning all three registers. But Adelle is a slab serif, so adopting it means a body-type shift away from Garamond's Renaissance humanism toward a warmer 20th-century editorial register. Paid. - **Brill (John Hudson for Brill Publishers) — serif + sans, no mono.** Designed specifically for humanities publishing; most voice-aligned for ARC's literary/scholarly register. Licensing: free for scholarly personal use; commercial requires purchase. Worth exploring with careful read of the license. **Reconsideration triggers for the specific families (not for Pattern 3 itself):** - Voice tension between Garamond and Plex becomes aesthetically costly in lived use - A specific future workstream (e.g., the typography application skill being built) makes the families' mismatch operationally visible - The steward completes appetite-assessment for font investment and decides a paid superfamily is worth the commitment - Open-source alternatives (Alegreya Sans, Ysabeau, iA Writer Mono) mature into production-readiness for ARC's specific needs **Naming for the record.** The steward's framing on 2026-04-18: *"Assigned roles to family seems logical to me and orienting (I am heavily myopic and now need progressives, so I tend to think of people who may need support visually — so serif for the body, sans for nav, and mono for apparatus seems to make things clear)."* This sentence is the canonical articulation of the Pattern 3 decision for ARC. ### Web typography spec coverage — see dedicated memory The link work of 2026-05-02 (§VII.c spec drafted; SCSS migrated; soft-tint tuning iterated) surfaced that *humble* typography surfaces had been left to browser-defaults plus ad-hoc accumulation. Same diagnostic likely applies elsewhere. Steward direction: *"the work needs to be bien-fait."* Tier 1 (lists · inline code + code blocks · headings hierarchy beyond H1), Tier 2 (smart-quote / em-dash conventions · numerals lining-vs-old-style · `::selection`), Tier 3 (`` / `` / `` / print stylesheet / `prefers-reduced-motion`), Tier 0 done, plus genuinely-don't-need items, all inventoried and ranked in: → `project-arc-web-typography-spec-coverage.md` This sits adjacent to Style Guide redevelopment but is distinct: the Style Guide is *internal vocabulary* (semantic post types, ornament classes, conventions); the web typography coverage is *element-level CSS spec* (what does a list / heading / code block render as). Both feed each other; should be read together when style-guide work resumes. ### §IV ornament retirement — content cleanup needed (2026-05-03) SCSS audit (2026-05-03) surfaced major spec/content mismatch: §IV explicitly retires the seven `.ornament.*` classes (.completion, .asterism, .marginalia, .lozenge, .gloss, .glimpse, .photo-essay), but content corpus has ~213 `.ornament.*` uses across markdown files. **Steward decision:** ornaments are retired per spec; *content* needs to be cleaned up, not just SCSS. Migration work (substantial): - Per-use audit of ~213 ornament instances → classify each into §IV-stated replacement (`class:` frontmatter / §II silence / future Pause/Close ornament) - Edit affected markdown files (~50-100 files, multiple uses each) - After content migration: SCSS `.ornament` rules retire from `_utilities.scss` - Update §IV to confirm retirement is operative not aspirational → `project-arc-ornament-spec-vs-content-mismatch-pending.md` — full details, decision-record, sequencing **Tied to:** `project-arc-pause-close-ornament-and-hr-audit-pending.md` — the Pause/Close ornament design that establishes what the .completion / piece-end uses migrate INTO. Realistic order: Pause/Close design first (so we know what content migrates to), then content audit + migration, then SCSS retirement. ### §IV Pause/Close ornament + hr audit — design needed (2026-05-03) Closely related to the ornament-retirement migration above. §IV describes Pause/Close ornaments functionally (paired-and-related cul-de-lampe at two scales) but the asset itself is yet to be designed. Steward direction: *"finis coronat opus"* — needs serious study. `
` content also needs audit (some legitimate Pause anchors, some incorrect carryover). Interim: three-dot `· · ·` per `_typography.scss hr` rule. → `project-arc-pause-close-ornament-and-hr-audit-pending.md` --- ## Chamber voices — archived 2026-04-18 The `content/chamber/voices/` directory (index.md + alexander/fusus/ngana-kulintja voice files) was archived to `docs/archive/chamber-voices/` 2026-04-18, authorized by the steward in the course of unblocking the Hakyll versioning infrastructure deployment. **Why:** a pre-existing dependency cycle in `content/chamber/voices/index.md` (the index's compile called `loadAll` on its own directory, triggering Hakyll's cycle detection despite the post-load filter). Flagged in `site.hs` as "TODO MODE-2: Re-implement voices with proper dependency resolution" — longstanding. It was blocking every full site build. **What was removed alongside:** voicesCtx in site.hs (dead code after branch removal), voices-nav links in `templates/chamber-canon.html` and `templates/chamber-deliberations.html`. **What remains operative:** the `voices_featured` field in `deliberationCtx` — deliberations still reference voices by name in their frontmatter; the text-level reference survives the archival. **Reinstatement condition (steward's note):** *"please note as it could be something to consider once the source aware chamber is fully implemented."* When the Chamber/BMF integration matures — i.e., once BMF recall can enumerate voice appearances across the corpus with proper source-awareness — the voice-as-first-class-document pattern can be reimplemented. At that point the index would be *generated* from a recall query, not globbed from markdown, which sidesteps the dependency cycle entirely. The archived files are reference material for that future work. Source-aware Chamber is downstream of L1 reliability stabilization and Chamber Phase 3 (currently parked). ## TODO: Reading Compass optical rebalancing (cosmetic, deferred) The 9-point Reading Compass nav uses per-glyph `font-size !important` rules in `scss/AldineXXI/_compass.scss:33-72` attempting to normalize perceived weight across the Unicode symbols (●○⟐◈◉⊙☷✎✍︎). Current values: Essays ● at 0.75rem, Observations ○ at 1.1rem (baseline), Chamber ⟐ at 1.35rem, Fragments ◈ at 0.8rem, Glimpses ◉ at 0.7rem, Photo Essays ⊙ at 0.85rem, Archive ☷ at 1.25rem, About ✍︎ at 1.3rem. **Known not-quite-working.** Steward observation 2026-04-18: *"the reading compass glyphs are meant to be adjusted to each other to be optically equivalent sizes, but I never got that working."* **Why it's hard.** The values were chosen once and never iterated against actual rendering at scale. Unicode symbols fall back to different system fonts depending on platform (Apple Emoji, Segoe UI Symbol, Noto Sans Symbols2, etc.), so a value that looks optically balanced on one reader's machine can look wrong on another's. The optical-size problem is real but platform-variable. **Possible paths to resolution** (not scoped here): 1. **Custom SVG glyphs** — replace Unicode symbols with in-house SVG icons designed together with proper optical equalization. Full control, platform-independent, but a design project in itself. 2. **Iteration with multi-device testing** — tune the font-size values against actual rendering on macOS, iOS, Windows, Android, common Linux. Cheaper than SVG but still ongoing maintenance as OS font rollouts change symbol rendering. 3. **Accept the platform variance** — document that compass symbols render differently across platforms and that the current values are best-effort, not pixel-perfect. Simplest. **Deferred** — cosmetic, not blocking anything, genuinely non-trivial. Worth revisiting alongside the style-guide session or when there's appetite for a glyph system. --- ## Mobile review list — 2026-04-18 evening (steward from mobile, Kai napping) Captured verbatim from the steward's list, prefixed `#ARC`. Not urgent. Mix of small mechanical fixes and genuinely philosophical questions. To be taken up in future sessions — likely alongside the style-guide session for the philosophical ones, and as quick-passes for the mechanical ones. **Mechanical / small:** - **Typography page needs updating** to reflect reality and established spec — *"the voices notably"* (with chamber voices archived 2026-04-18, any references on the typography page to the voices system are now stale) - **Colophon — *The Formation* missing a context sentence** (cf the *Guides* section which has one) — the *Formation* section added 2026-04-17 has entries but no introductory context sentence - **Colophon mini nav: no link to AldineXXI** — the in-page navigation at colophon footer should link to the AldineXXI typography page - **Remove Patterns from about pages** — pure pedagogy, probably belongs on the personal site (eventual `davidglidden.eu`). To be decided upon properly later; for now, remove. *Steward's framing: "To be decided upon, but for now, let's remove."* - **Glimpse on mobile not edge-to-edge** — responsive CSS issue; class expects full-bleed per spec, not rendering edge-to-edge on mobile viewport - **Apparatus: location should always be included** — currently conditional (`$if(location)$` in post.html). Also open question: *italics?* (the `.dateline-location` class already has `font-style: italic`, but the question may be whether that's the right choice or should be revisited) **Philosophical / needs thought:** - **About enfilade: does it need updating?** — *"What stays? What goes?"* With the decoupling horizon in view (ARC → `animalrationiscapax.org`, personal site → `davidglidden.eu`), the about-page enfilade navigation's scope may shift; worth auditing now regardless. - **Reframe (2026-04-18 evening, verbatim steward):** *"the enfilade acts as both access and map."* The enfilade declares the about-territory to the reader. - **Architecture clarified (after multiple misreads):** The about-enfilade is **inline on each about-constellation page** (not a template partial). Each page maintains its own `