Film / narrative site
The Fractured Archive
An episodic film site with a twelve-entry record and a release gate enforced on the server, not in the page. Ten episodes published, about 29 minutes of video live.
thefracturedarchive.orgA one-person studio. I write the world, produce the film, set the book, and build the system that ships it. Everything on this page is live and clickable, or it isn’t on this page.
12
Episodes produced
391
Clips cut down to twelve
129
Print-ready plates, 7 volumes
7
Applications live in public
Five public sites, all verified responding on 22 August 2026. Click any of them — that is the whole point of listing them.
Film / narrative site
An episodic film site with a twelve-entry record and a release gate enforced on the server, not in the page. Ten episodes published, about 29 minutes of video live.
thefracturedarchive.orgInteractive web artifact
A grimoire in four movements — Door, Registry, Oracle, Rite. 32 relics and 60 transmissions, 21 of 21 tests passing, with a published threat model that includes what it does not cover.
registry.thefracturedarchive.orgStorefront
A catalogue of thirteen physical marks and seven colouring volumes, each with its own product photography and checkout.
reliquary.thefracturedarchive.orgOperations software
Equipment inventory built for the photography company I work at. In use by seven people since June: 527 items, 40 recorded handoffs, 39 reservations, QR check-in and out, and reservations that refuse to double-book.
loadout.voxverra.comAutomation
A digest engine I built and run for myself: it watches its own sources, runs its own tests and takes its own backups without being asked. Not a product and not for sale — it is here because the machinery is the point.
curio.voxverra.comThe next one
Whatever you are making, it can be on this list — live, linkable and yours.
Start a project01
Short films and episodic series taken from a source text through shot planning, production, grade, narration and a delivery pipeline that actually publishes them — renditions, posters, cache-busting and all.
12 episodes shipped
02
A universe with rules that hold: canon, characters, a written voice, and the continuity machinery to keep a long work from drifting. Built to be extended by other people without breaking.
133 canon entries
03
Manuscript or footage to press-ready files. Real production specs, not approximations — trim, bleed, spine width from the printer’s own formula, bilevel interiors, blank versos where ink would bleed through.
129 plates, 7 volumes
04
Colour, type, grid, spacing, states and the actual screens — delivered as a working stylesheet and live pages, not a slide deck. No framework to go out of date, and every decision written down for whoever comes next.
45 interactive pieces, 45 territories
05
One tool for the thing your team currently runs on a spreadsheet, a whiteboard and a group chat. Roles and permissions, scan in and out, reservations that refuse double-bookings, email that actually sends, nightly backups.
In daily use by 7 people
Stage 1
Thirty minutes, free. What you are making, what is in the way, and whether I am the right person. Sometimes the answer is no.
Stage 2
A document you approve before anything gets made. If I scoped it wrong, I re-scope once at no cost.
Stage 3
You get a working URL or a first cut early, not a reveal at the end. Course corrections are cheap in week one and expensive in week eight.
Stage 4
Files, source and a written guide, in your accounts and your name. Nothing important stays on my machine.
Stage 5
Fixes after handover are included, not a line item. After that there is an optional care plan if you want one.
Three hundred and ninety-one clips were cut down to twelve episodes. The first build of the colouring plates came out scratchy and broken, so it was thrown away and rebuilt from scratch rather than shipped. Every stock narrator voice was rejected in favour of my own. A composition signature collapsed 314 candidate pages to 230 unique ones, catching 47 near-identical repeats that a perceptual hash had missed.
The work is in the deciding and the discarding. Making a thing is the easy half. The job is knowing which one is right, how many attempts that takes, and being willing to throw away the first finished version when it is not good enough.
The same applies to the software. No frameworks, no build step, no dependency that can break on someone else’s release schedule — every application listed above is written against the standard library and will still run in five years.
Tell me what you are making. I reply to everything within one business day, and I will tell you plainly if it is not something I should take on.