Skip to content

Videoform: rules in a taste line's shape, and where to start - #208

Open
shinyobjectz wants to merge 1 commit into
mainfrom
fingerprint-one
Open

shinyobjectz wants to merge 1 commit into
mainfrom
fingerprint-one

Conversation

@shinyobjectz

@shinyobjectz shinyobjectz commented Nov 20, 2026 •

Copy link
Copy Markdown
Owner

This PR is now a small change on top of the videoform plan as Zach wrote it (#207, merged).

The six-way merge this PR used to hold is gone: no rules.ts, no moving taste files, no folding raea judge in. It's kept on fingerprint-one-full in case it's wanted later.

What changes

  • VF8: rules in a taste line's shape. Controllers are being removed (PJ9: controllers are gone; a standing rule is a taste line #209), so a videoform's mined rules go in its own rules. Each rule has a taste line's shape: words, a check, its values and a level.
  • The check hook. raea check PROJECT holds a piece to the rules of the fingerprint that judges it (which), beside the person's, the workspace's and the project's taste. It's a 15-line hook in apps/direction/src/check.rs.
  • "Start here" in docs/videoform.feature:
    • the order of work, with VF4 plus VF12 as the gate before VF1b's platform pages and VF9-VF11;
    • the new crates;
    • the three small hooks into existing code (the videoform field, raea context, raea check);
    • the bench.
  • docs/fingerprint.md decision 6: Pro and Studio's monthly pool of clef checks and content intelligence, which can find a videoform's references.

It's docs only: 3 files, 40 lines added and 14 removed.

@wackyza

wackyza commented Nov 20, 2026 •

Copy link
Copy Markdown
Collaborator

Checked against #207 sentence by sentence and task by task. Nothing from the videoform plan was lost. All 17 VF tasks are here with their file plans, and the only changes are the two you listed (mined rules in rules.ts; raea fingerprint in place of raea videoform). report folding into raea fingerprint report keeps VF5 working. The consolidation makes sense to me. Three notes:

1. Keep raea check fast (FP15e). Putting the axes' flags into every raea check would mean SigLIP over every frame. The FP3a/FP1 numbers put that at about 17 s for a 90 s render on ort's CPU path (47 ms a frame).

Why that matters: we measured where a check's time goes for #205. On 56 s exports it took 13 to 20 s, almost all of it four separate full decodes of the video (flashing, freezes, flash frames, blotching), about 3.7 s each. Running them in parallel didn't help: four together took 15.8 s against 3.7 s alone, because decoding already uses all 8 cores. Only decoding once helped, which is what #205 did: the same findings byte for byte in 7 to 8 s. A check runs after every render, so it's paid every round. Per-frame SigLIP on every check would cost more than the whole check does now.

So the flags should come from cached columns (Fingerprint already caches them by content hash), and SigLIP should run only on shots that changed since the last check, as VF11 already says for videoform checks. Then the one check stays as fast as it is.

2. Gate on consistency, not on how many references there are (VF8b, FP15e). VF8b, from my plan, says that under five references nothing gates. I'd change that. The encoders are frozen: a videoform learns from how consistent its references are, not from training a model, and 3 to 5 very similar videos is enough. Some of the people we're talking to already edit to a formula by hand, so their references will agree tightly, and that agreement is the signal.

  • A rule or range that holds tightly across every reference gates, with as few as 3.
  • A trait that varies stays free, however many references there are.
  • Under 5, the leave-one-out score is reported as low-confidence instead of switching gating off.

FP15e should also give rules a level from their strength, as it does for axes, so a weakly held mined rule is a warning or a note and can't fail a render on something another render won't move.

3. Rule precedence by origin (FP15d). "The nearer one wins" settles two rules on the same check at different levels. At the same level, I'd settle it by origin: a rule the person wrote, then one kept from a note, then one mined from references. That follows the floor's "the person wins".

4. An approved edit becomes a reference (FP15f, VF8c). FP15f routes every label into the fingerprint that judged the piece, which covers learning from notes and picks. What isn't written anywhere: when the person approves a cut made to a videoform, it should be added as a reference, read once like any other (VF3), so the videoform's ranges, shot kinds, structure and rules are remade from its rows (VF8c, in seconds). Then a videoform improves with every approved project, not only from corrections.

This fits the three layers already in the plan, which is also why "frozen encoders" in point 2 isn't a limit:

  • the reference statistics (the videoform), which work from 3 to 5 videos;
  • the small heads trained from feedback in milliseconds (FP2c);
  • the encoder adapter only for a trait still stuck after many labels (FP10).

Approved edits feed the first layer, and their notes and picks feed the second.

Controllers are gone (projects.feature PJ9, #209), so a videoform's mined rules go in its own `rules`, each in a taste
line's shape (words, a check, its values, a level), and `raea check PROJECT` holds a piece to the rules of the
fingerprint that judges it (VF8, a 15-line hook in apps/direction check.rs). docs/videoform.feature gains a short
"Start here": the order (VF1, VF2, VF3, then VF4 with VF12 as the gate before VF1b and VF9-VF11), the new crates, the
three hooks into what exists and the bench. docs/fingerprint.md records decision 6: Pro and Studio's monthly pool of
clef checks and content intelligence.
@shinyobjectz shinyobjectz changed the title One fingerprint: taste, rules, axes, ranges, a rubric and a videoform (spec, FP15) Videoform: rules in a taste line's shape, and where to start Nov 20, 2026
@shinyobjectz
shinyobjectz requested a review from wackyza November 20, 2026 22:10
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants