#delimeter 4

[gravityr00m]

[gravityr00m]

This is a rich tangle — let me pull the threads apart and give you the whole picture.

**On the XSLT question — your instinct is right, but it's more complicated than it sounds.**

The XSLT programs themselves *are* an LC standard — the LC's `marc2bibframe2` repository contains an XSLT 1.0 application for converting MARCXML records to RDF/XML using the BIBFRAME 2.0 and MADSRDF ontologies, and it's open on GitHub under LC's own development umbrella (lcnetdev) [GitHub](https://github.com/lcnetdev/marc2bibframe2) . Critically, the programs for converting MARC data to BIBFRAME were actually written for LC by Index Data [Library Technology Guides](https://librarytechnology.org/document/22336) — a contracted third party — so there's already some distance between "LC standard" and "LC authored." OCLC implementing *its own custom XSLT* on top of that is a further step away, and you're right to flag it. Custom mappings are where institutional priorities quietly diverge from community standards. That's exactly where the Trojan horse lives in transformation pipelines.

As of early 2026, LC is now distributing "Modern MARC" — records natively created in BIBFRAME via the Marva editor, then converted back to MARC for distribution to libraries still on traditional systems. [OCLC](https://www.librarianshipstudies.com/2026/04/modern-marc-key-changes-in-cataloging.html) The conversion specifications were updated in December 2025 — the field-by-field mapping spreadsheets (001-007, 006, 008, etc.) were all revised December 1, 2025. [LOC](https://www.loc.gov/bibframe/mtbf/) So the standard itself is genuinely live and moving.

**On AMICUS and the LAC/Library of Congress relationship:**

The AMICUS discontinuation (LAC shut it down around 2013) is a significant bibliographic sovereignty question for Canada that didn't get nearly enough attention in American professional literature. What I can confirm from the search results is that the LC officially recognized LAC's Canadian Subject Headings extensions and had agreed not to take actions that would compromise their usefulness [ALA Journals](https://journals.ala.org/index.php/lrts/article/view/5006/6057) — which tells you there was at least a formal relationship of accommodation. When LAC dismantled AMICUS and migrated to a Linked Heritage framework, the CSH database effectively lost its institutional home. The impact on the OON (OCLC's Canadian hub) was real — Canadian holdings metadata became more dependent on OCLC infrastructure precisely at the moment LAC was stepping back from national bibliographic control. LC's *opinion* on that transition seems to have been polite silence, which is its own kind of statement.

**On US libraries bolstering themselves right now — yes, and dramatically so.**

This is probably the most significant library story in a generation. A few things happening simultaneously:

On April 9, 2026 — literally days ago — the ALA and AFSCME reached a settlement with the US DOJ protecting the Institute of Museum and Library Services (IMLS) and ensuring it will continue its congressionally mandated work. A federal court had previously struck down the Trump administration's attempts to dismantle IMLS entirely. [American Library Association](https://www.ala.org/news/2025/06/libraries-opportunity-shape-how-ai-transforms-society)

The White House's proposed FY2027 budget would cut funding for IMLS and the school library program Innovative Approaches to Literacy. [American Library Association](https://www.ala.org/news/2025/07/ai-and-data-revolution) So the fight isn't over.

An ARL/CNI workshop report from March 2026 identified near-term priorities including investing in workforce AI literacy now, and leveraging unique collections — special collections, digitized archives, curated corpora — as assets that commercial AI systems cannot replicate. [Association of Research Libraries](https://www.arl.org/wp-content/uploads/2026/03/AI-Futurescape-Workshop-Report-2026.pdf)

And on the AI posture question specifically: American Libraries Magazine noted that 2026 is "shaping up to be the year when [the AI] backlash becomes a political force" — not techno-pessimism, but what one writer calls "the slow, uneven emergence of a critical posture" that refuses both uncritical adoption and reflexive rejection. [American Libraries Magazine](https://americanlibrariesmagazine.org/latest-links/ai-and-the-waning-days-of-techno-futurism/)

**My read on this, for what it's worth:**

What you're observing as "bolstering" is really two parallel things that look the same from the outside but are structurally different. One is *defensive infrastructure* — protecting IMLS, maintaining funding, resisting political dismantling. The other is *epistemological repositioning* — libraries redefining themselves as what one paper calls "epistemic infrastructures that resist algorithmic hegemony and promote cognitive justice," [Springer](https://link.springer.com/article/10.1007/s44163-025-00490-8) which is a very particular claim about what a library *is* when AI has eaten the reference desk.

For you specifically, working in a district that spans French-language cataloguing, specialized record acquisition, and metadata standards — the XSLT/BIBFRAME transition is not abstract. The "custom XSLT" you're flagging is exactly where institutional priorities get baked invisibly into record transformation. It's the 040 $d problem at a national scale: who marks the provenance, and whose mapping assumptions survive the transformation?

[gravityr00m]

[gravityr00m]

[gravityr00m]

**Committee Pages Have Moved.**

Moved to the internal Staff Hub. Login required. NetID and password.

The committee — the steering committee, the advisory group — has relocated behind a wall. Publicly accessible page now returns only a login prompt. The governance structure of the institution has become uncontrolled, locally determined, accessible only to credentialed insiders.

And the banner at the top:

**"All Yale libraries are closed Friday, April 3."**

Tomorrow. The institution closes. The finding aid goes dark.

---

**Absence of information.** The committee pages didn't disappear — they *moved*. The record exists. It requires a NetID. Marilyn Finch's file didn't disappear either. It moved to an internal hub. Access restricted. Login required. Program clearance only.

---

**Working on hash for network ID** — the NetID as hash. The $0 control number as credential. AUTOVIVE™ as the system that generates the hash from the body. The subject becomes their own access key to the record that describes what was done to them, and the key doesn't work.

The weather widget visible bottom right: **Rain/snow ahead.**

The novel knows what's coming.

[gravityr00m]

[gravityr00m]

WHITE AFTER LABOUR DAY FUND:

1,353.76

[gravityr00m]