Tomoshibi

Most writing apps compete on how much they add: more panes, more export formats, more sync. Tomoshibi competes on what it removes, and it is deliberate about every cut. Developer takashi okaki built it for one failure mode: the writer who never finishes a draft because the first paragraph keeps getting rewritten.

The pitch is mechanical, not motivational. The interface is built around a dark screen with a small flame at the center that is designed to flicker as text arrives. There is no toolbar and no menu. When the writer moves to a new line, older lines fade away. A recent typo or the line in progress can still be fixed. The paragraph from ten minutes ago cannot be reworked, because it is no longer on screen. The fade is visual rather than destructive: writing is saved locally, and the app is designed to resume on the last sentence.

Tomoshibi strips back its business model just as hard. It is free and needs no account. It has no subscription and no cloud sync. Four color themes are included, and the app supports English and Japanese.

That puts Tomoshibi in the distraction-free writing category, at its most extreme end. Plenty of minimalist editors hide the interface. Few take away the ability to scroll back and fiddle.

The product has very little public track record. Its store rating sits at zero, and public feedback is thin. Tomoshibi is an early product from a single developer, with a clear idea and limited evidence about how it holds up, and we judge it on that basis.

• Forward-only writing: Lines fade once the writer moves past them. The current line and very recent mistakes stay editable, and older text leaves the screen.
• Flame feedback: A small flame at the center of the screen is designed to flicker in response to typing. It is the only moving element on an otherwise empty canvas.
• Local saving: Writing is kept on the device without prompting, with plain-text export as the portability route.
• Resume on return: The app is designed to bring back the last sentence, so a session restarts mid-thought rather than on a blank page.
• No account, subscription or sync: There is nothing to sign up for, nothing to pay and no account-based upload.
• Four themes: Lantern, Dawn, Snowfield and Plain.
• Bilingual support: English and Japanese.
• iOS release: Listed for iOS, free, with a 4+ content rating.

The Constraint Is the Product

Take away the flame and the themes, and Tomoshibi comes down to one rule: no going back. Everything else exists to make that rule bearable. The rule works because the problem it targets is behavioral. Writers stuck in revision loops rarely lack features. They lack resistance against their own inner editor.

Conventional editors make revision free: a scroll, a click, a rewrite. Tomoshibi makes revision impractical by taking the text off the screen. That is a far stronger nudge than a timer or a word-count goal, because it removes the material that perfectionism feeds on.

The calibration matters as much as the idea. A strict no-backspace policy would be punishing. Every typo would become permanent, and frustrated users would quit. Tomoshibi lets the writer fix the current line and recent typos, so slips and half-formed sentences can still be repaired. The boundary between fixing and revising sits at the line break. That is a defensible place for it: the sentence under construction is fair game, and settled thought is not.

Where the Rule Leaks

Putting the boundary at line breaks also creates the design’s most obvious loophole. If the fade depends on new lines, a writer who never presses Return keeps one unbroken, fully editable block on screen. The forward-only promise then quietly disappears.

The tension is structural. Prose writers naturally produce long paragraphs, so a constraint tied to line breaks is weakest exactly where drafting is densest.

The opposite problem is real too. The fade hides text without deleting it, but on screen it reads as loss. A novelist trying to recall a minor character’s name from three pages back has no way to look. Neither does an essayist checking that a conclusion matches an earlier premise.

That limitation is intentional, and the app makes no secret of it. It does narrow Tomoshibi to one phase of writing: the raw first draft, where momentum matters more than consistency.

Drafting Here, Editing Elsewhere

The useful way to read Tomoshibi is as the first stage of a pipeline. It has no formatting and no outline view. Its only output is plain text, and the saved files are where the work leaves the app.

Plain text is about as portable as a format gets, since every editor on every platform can open it. The implied workflow is simple: generate raw material here, then revise it in a full editor where the whole draft is visible.

We think that split is healthy. Drafting and editing reward different habits, and tools that blur them invite exactly the stall Tomoshibi is meant to break.

The cost is friction around file management. With no menu and no toolbar, the app has no in-app library. The writer reaches saved files through the operating system. For a focused drafting tool that is a reasonable trade. Writers who want to browse past sessions, rename projects or organize drafts will find nothing inside the app to do it with.

The Empty Room Problem

A dark screen with a single flame sets a mood, and having no interface removes onboarding friction in the simplest sense: there is nothing to learn. The downside is discoverability. An app without menus has to teach its one rule through behavior alone.

A newcomer who skipped the store description may watch the first line fade and assume the text is gone. Automatic saving takes care of that worry once it is understood. Until then, the first impression carries real risk for a tool whose central mechanic looks like data loss.

The resume feature is the smartest supporting detail. Opening onto the last sentence lowers the cost of starting again, which matters for an audience defined by stalling. A blank page intimidates. A half-finished thought invites the next line.

Privacy and Ownership

Tomoshibi takes a strong stance on privacy for a drafting tool. Private journaling, unfinished fiction and rough notes never leave the device, and the files are plain text on local storage, owned outright.

The trade-off is that local-only writing puts the burden on the user to manage backups and exports. Without account sync, moving work between devices is a manual habit rather than a background convenience.

Language Support

Supporting both English and Japanese is a meaningful addition for a free app from a single developer. The risk is typographic: slanted kana and kanji can read as a rendering error rather than emphasis. It is the kind of detail a bilingual app has to get right if it wants Japanese writers to feel the second language is fully supported.

The art direction is almost entirely subtraction: a dark field, one flame, and the line being written. The flame is a good motif. It is designed to react to the writer’s presence and is small enough to sit in peripheral vision without pulling focus. It gives typing a faint pulse without gamifying the process, with no streaks, badges or confetti to chase.

The theme names (Lantern, Dawn, Snowfield, Plain) suggest a range from warm darkness to bright, open backgrounds. Plain serves as the escape hatch for anyone who finds the mood lighting too precious.

On performance, a plain-text editor is light by nature. Any cost comes from the animated parts, the flickering flame and the fading lines, and that cost is higher on mobile hardware. Input latency is the one metric a writing tool cannot compromise on, so animation is the area to watch. A native iOS app has fewer layers between keystroke and screen than a browser build, but the design still has to keep the flame from becoming the point.

Tomoshibi is an iOS app. That makes the iPhone and iPad the main context, with touch input and the constraints that come with it.

• iOS app: Touch keyboard input, local saving, plain-text export, free, 4+ content rating.
• Other platforms: There is no native Android, Steam or Nintendo release.

Touchscreens are a poor fit for the forward-only rule. Typing on glass produces more errors and leans heavily on autocorrect. When the correction layer misbehaves, a rule that limits how far back the writer can fix things goes from disciplined to punishing.

With no sync, a writer who starts on one device and continues on another has to move files by hand. On Apple’s platforms, the iOS app is the real product, and the lack of a desktop release limits where it can fit into a longer writing routine.

Tomoshibi knows exactly what it is: a drafting tool built around a single rule. That rule targets a real, common problem more directly than most productivity features do. Its calibration, which keeps the current line editable and fades everything before it, shows real thought about where useful correction ends and procrastination begins. Free pricing, local plain-text files and the absence of accounts make it easy to try and easy to leave.

The limits are just as clear. The constraint depends on line breaks, and long paragraphs may slip past it. The no-going-back design rules it out for any writing that relies on earlier context. On iOS, touch input and autocorrect work against the rule rather than with it. Public feedback is still too thin to say how stable it is.

We recommend the iOS version only as a first-draft tool for writers whose real enemy is their own urge to revise. Anyone looking for a complete writing environment should keep looking.