Loading the documentation…
Prepared views
Named starting points — New Releases, Short Stories, whatever fits the catalog — composed in the Librarian, one click for the reader.
jpl · jay pscot library
Publish your writing
as a living library.
A self-hosted, tag-driven library for authors — books, novellas, short stories, and articles organized by ideas rather than shelves. One catalog, prepared views, reader-controlled refinement by tag, author, and series — and a Librarian's desk behind it. Framework-free PHP + SQLite; no Composer, no Node, no build step.
$ curl -O https://smisco.biz/jpl/jpl.zip
the proof
It runs its author's own catalog.
pscot.com is not a demo rig — it is this library, live, serving Jay Pscot's books, stories, and articles to readers. Every screenshot on this page is that instance, doing its day job.
the reader's side
Start with a view. Refine the results.
Prepared views are starting points, not silos. Tags, authors, and series stack: every selection narrows the shelf, every removal widens it, and text search runs across titles, subtitles, summaries, author names, and series names. Refinements survive a change of view, update without a page reload, and live in the URL — so any shelf a reader builds is a link they can send, one click away beside the results heading.
behind the desk
Every library needs a Librarian.
The admin side lives at /desk and speaks library. The Circulation Desk is the dashboard: catalog totals, authors, series, tags, views, visitors, Reads, Previews, Downloads, and Amazon hits — all first-party, no third-party analytics. Authors and series are records, not free text: each has its own page, and selecting a row lists every catalog entry filed under it. The catalog itself filters by author, series, and status, or searches across all of them. Entries carry covers, grouped tags with quick-add, a formatted summary, and two files: the public document — the exact, complete file visitors receive (PDF, EPUB, Markdown, HTML, Word, and more, up to 50 MB) — and an optional Librarian-only file of any type that never gets a public route. Per work, the Librarian decides what visitors may do: open the public document (labeled Preview or Read), download it, or follow an Amazon link — each independent.
what you get
The whole library, in one zip.
Tag refinement
Grouped tags (format, genre, subject) with quick-add; every tag narrows, every removal widens.
Authors & series
First-class records, not free text. A work can belong to several of either; both filter the public library and both get their own Librarian page.
Shareable search
Tags, authors, series, and text search across titles, summaries, and names all live in the URL; a link icon and Copy URL beside the results hand over the exact shelf.
Librarian-only files
Each entry can carry a private attachment — a full manuscript, a review draft, any file type — stored outside the public site and opened only at the Desk.
Work pages
Covers front and back, subtitles, series, bylines, summaries — each work with its own page and slug.
Formatted summaries
A small, enforced subset of HTML — emphasis, quotes, rules, code. Attributes stripped, unknown tags removed, sanitized on save and on display.
Whole documents, own origin
Visitors get the exact file the author prepared, never trimmed, served from a separate reader host — so an HTML reader's scripts and saved settings work, and the Librarian's session stays out of reach.
Amazon linking
Commercial works link out to Amazon — independent of read and download permissions.
First-party analytics
Reads, previews, downloads, visitors — hashed visitor tokens, no IPs stored, nothing sent to third parties.
Private data, private
SQLite database, Librarian-only files, covers, and backups live outside the webroot; public documents sit behind a reader route that never exposes their storage paths.
Type scale & PWA
One setting sizes the whole library and desk, 80–140%. Installable as an app, with the library's own mark and colors.
requirements & install
PHP 8.2, SQLite, and three folders.
Requirements
- PHP 8.2+ with pdo_sqlite, fileinfo, json
- Apache with .htaccess support, or equivalent server rules
- HTTPS on three hosts, each with its own site folder: the main site, a docs. subdomain for the isolated document reader, and a desk. subdomain that simply redirects to /desk/
- Explicitly no Composer, PHP framework, JS framework, Node, build system, or database server
Private data (the SQLite database, Librarian-only files, covers, backups) lives beside — not inside — the webroot; public documents live in the reader host's own tree. The README covers the production layout, a fresh install, and the upgrade path; upgrades copy over the existing tree and leave the catalog, tags, authors, series, files, and analytics intact.
Install
# three site folders, HTTPS on each: # main → the library + /desk/ # docs. → the isolated document reader # desk. → a redirect to /desk/ # extracts to pscot-site/ — place it as the site tree: $ php bin/install.php # also sets up the reader host $ php bin/check.php # sign in at /desk/ and start cataloging
why it's free
It costs nothing. Tips are appreciated.
The Jay Pscot Library is free — the whole of it, MIT-licensed. There is no subscription, no paid tier holding features back, and no advertising. Download it and it is simply yours to run.
If it gives your writing a home, you can leave a tip. It is a thank-you for the tool, never an obligation — and it is always appreciated.
rendered live · from README.md
Straight from the zip.
The document below is the current distribution's own README — requirements, the upgrade path, and what changed in this release — fetched and rendered in your browser by md-mini.js — a sibling smisco tool. The zip also ships RELEASE.md and DESIGN.md alongside it.