Oryne

Oryne is an iPhone app for people whose best ideas get lost in a notes list. Say or type a thought and let it go. The phone names it, drifts it next to related thoughts, and brings it back on a day it might matter.

Every notes app promises to keep your thoughts organized, and every one breaks the promise the same way: not by losing them, but by burying them. Neatly, chronologically, in a list you never scroll back through.

Lists optimize for putting things in. Minds optimize for things coming back. A thought from three months ago surfaces because something today rhymes with it, not because you scrolled to March.

The list is honest about storage and dishonest about memory. The problem was never the list, it was the monopoly. So the design had to start from how a thought comes back. Every tool a creative person already keeps inspiration in has an answer to that, and they are not the same answer.

One question, put to each tool a creative person already keeps inspiration in: what happens to a thing after you save it, and whether it ever finds you again. I asked it in February, three and a half months before the first commit.

Three capabilities settle the answer. Capture has to be fast enough that you do it at all. The saved thing has to live somewhere you can look around in, not just scroll. And something has to return it to you without being asked.

Tool

Fast capture

A canvas

Resurfacing

Pinterest

One tap

An endless grid

Other people’s ideas

Milanote

Mobile quick note

Boards you arrange

Only what you open

mymind

One click

A visual grid

Serendipity

Freeform

Place it yourself

An infinite canvas

Only what you placed

Notion

Quick add, then file

Pages and databases

A search box

Oryne

Two seconds, by voice

Currents that drift

A thought a day

Each one breaks the chain somewhere. mymind comes closest, and it collects what you find rather than what you think. Nothing was doing all three for a thought of your own, and that is the gap the idea was built in.

The gap was a tool that catches a thought, gives it a place, and brings it back. Water does all three without a filing system: things drift, related ones gather, and what sank comes up again on its own. So the mind became an ocean, and a thought became something you let go of rather than something you put away.

That only helps if the metaphor makes decisions, so it was held to one rule: every interaction has to be explainable as how water works, never as how apps work. Capture is a thought falling into the ocean. Categorization is a thought flowing into a current, and that is the literal copy in the app, because “assign to category” would have been the list sneaking back in through the language.

A field that feels alive is illegible by default. So the Ocean is for encountering, the Library is for finding, and neither pretends to be the other. A metaphor that settles questions this large should not live in one person's head, so it was written down as rules.

Two days into the native build I wrote the metaphor down as eight principles and put the file in the repository next to the code. Each one ends in a question, and the question is the whole point: a review can cite “violates Trust, confirmation before verification” instead of “feels wrong,” and if the answer is yes, the change does not ship. Five belong to one of the four places. Three hold everywhere.

A principle without a test is a mood. These eight were written for the third version of this idea. The two versions before it are why there was anything to write.

The metaphor arrived before the product did. It went through a web prototype, a set of wireframes, and a first native build before the Ocean looked like water, and each version kept something the next one still has.

01

Lovable, the first version

A Web Prototype Called InspireOcean

Cards on a dotted canvas, a list view one keystroke away, and a Rediscover panel that surfaced older ideas not opened recently.

What survived

The three verbs were already here: capture, browse spatially, and let old ideas come back on their own.

What changed

The canvas was a whiteboard, not water. Cards sat wherever they were dropped, nothing moved, and nothing was related to anything.

02

Claude Design, May 24

Wireframes That Argued With Themselves

Eight flows as storyboard rows, and under each hard question a row of side-by-side alternatives. Navigation, the Ask surface, the Library, and the capture entry points shipped as drawn. The three rows below did not.

What survived

Four tabs named for the four verbs, Capture, Ocean, Ask, and Library, and long press as the canonical gesture: chosen there for branching, kept in the shipped app for kinship.

What changed

The Ocean Field question was never settled by picking one. The build went Cluster first, on day two, and drift arrived in July. The shipped Ocean is both: currents that move.

Three Rows That Did Not Ship as Drawn

03

SwiftUI, from June 9

Version 1: The First Build

By the end of day one the four tabs existed. The Ocean was a handful of gray orbs on a dark field, the Library a list, Whisper a mic and a timer, and Ask a chat with four lenses. Fast Capture and the widgets followed on day two.

What survived

The skeleton shipped as it was: the four tabs, Thought and Whisper capture, Grow a Branch, Ask answering only from the water.

What changed

Orbs became glass bodies that drift in currents and light up their relatives on a long press. The Library became a waterfall you can sort by meaning. Ask kept its lenses and gained a floor under which it admits it has nothing.

Version 1 Against Shipped

Ocean

Library

What survived all three was the skeleton: four verbs, four tabs, and a long press.

Shipping did not end the iteration, it changed the clock. I had never written production Swift. Twenty-three days after the first commit Oryne was on the App Store, and five more releases followed within two weeks, each closing something the last had left open.

The agent wrote the code. I designed the review, because when the author and the reviewer share the same blind spots you do not have a review. You have a rubber stamp with extra steps.

Speed was the agent's job. Judgment stayed mine, and the gates exist so that nothing irreversible ever happens at machine speed.

One thought, start to finish, and the decisions underneath. Every screen is the shipped app.

Act One: You Catch It

Capture is built for the moment a thought arrives, before you have judged it.

01

Catch

02

Release

Act Two: The Ocean Works

Everything you never get around to happens on the device after you let go. Then, one day, it comes back.

03

Understand

04

Drift

05

Gather

06

Return

When an effect you love is fighting the information it carries, the effect loses.

Act Three: When You Go Looking

Rediscovery is a rhythm, not a search. When you do go looking, there is a place for that.

07

Find

08

Ask

Act Four: Around the Loop

Four more features, each built to keep the Ocean simple.

09

Fast Capture

10

Widgets

11

Grow a Branch

12

Show Related

No new gestures on the Ocean, ever.

Capture in under two seconds from anywhere. If a change would make someone hesitate before capturing, the answer is no. And what you are willing to capture in the first place depends entirely on where you believe it goes.

An inspiration tool only works if you will feed it your half-formed, 2 a.m. thoughts, so nothing you capture transits a server I can read.

On-device models are smaller than the cloud's, and classification, kinship, and resurfacing were all designed around what local intelligence can actually do rather than what a server could have done. I would make the same trade again.

On-device is not a spec-sheet line. It is the precondition for the honesty the whole product depends on.

Nearly every choice trades control for encounter: browsing over managing, drift over sorting, restraint over reach. That is the trade I keep wanting to make, and probably the most honest description of the kind of designer I am.