Familiar by design is not a better world — typographic card

Familiar by design is not a better world: open source and intellectual surrender

I am writing this in early October 2026 with more doubts than certainty. For the first time in open source, I feel something chilling: that we could have surrendered our intellectual strength to the big tech giants. I could be wrong. I hope I am. I do not like this feeling.

I am a developer, a GIMP user, not a Photoshop daily driver. I do not crave a tool whose selling point is that I will not have to learn anything new — “no retraining” because the menus already match Adobe. That promise does not pull me; I am fine living outside Photoshop muscle memory. What I keep hearing — suite clones, Rust rewrites, jokes about how Adobe will react — lands as a battleground on their ground. That may be clever. I am not sure it is liberation.

This is not a hit piece on the people shipping ArtCraft apps, and it is not legal advice. Clean-room claims, trade dress, licences, and patents are being argued elsewhere. I am stuck on something smaller and sharper: the product definition itself — and on that cold feeling that our imagination started from their end product, not from ours.

The slogan that gives the game away

On the PhotoCraft page the headline is: “The image editor you already know how to use.” Next to it, Familiar by design: menus, shortcuts, panels, and tools “your hands already know,” productive on day one, no retraining.

What can that mean, in this market, if not: like Adobe Photoshop? The page does not need to print the brand in the H1. The promise does the pointing.

It is not only Photoshop. On CADCraft the body says it without soft focus: “The command line, object snaps, layers, dimensions, hatches, blocks and drawings you already know.” That is AutoCAD’s grammar sold as the feature. Title lines may say “reimagined in pure Rust”; the sell is that nothing in your hands has to change. New substrate. Same map.

I read that as intellectual surrender. Not because interchange formats are evil — PSD and DXF matter — but because the story of the project becomes: we meet the incumbent on the incumbent’s face (for whom this is not clear enough: what I mean is we challenge Adobe by rebuilding Adobe’s visible product — menus, shortcuts, layout — instead of starting from a different idea of the tool). First blood goes to whoever already defined what “knowing how to use the tool” means.

Union air, not the ladder

If I imagine developers — or a broader commons — putting scarce effort into building something new, I imagine the output to look different from what already owns the market. If you “simply” give me the same world written in another language, the unionist in me is not happy.

“Familiar by design” is something I would never put on a picket sign to help workers. The hell with familiar. People want fresh air, not a prettier ladder to the same top. John Lennon’s Working Class Hero still cuts: keep you doped with religion, sex and TV — and there is still the ladder, if you want it. Familiarity can be a kindness. It can also be containment.

Meeting Adobe on Adobe’s ground may feel like a jab. Voices celebrating “how would Adobe react if someone clones the suite?” sound like battlefield tourism to me. Challenging a landlord by rebuilding the landlord’s floor plan brick-for-brick is a strange idea of freedom. I may be wrong. I keep being told I am missing the joke, or the file-compatibility win, or the subscription anger. Those are real. They do not dissolve the slogan.

Open source used to mean different weather

Let me be clear about what I am not against. A serious Photoshop alternative for Linux is a great idea to me — in the same family of intention as GIMP. GIMP’s whole idea is good. I use it sparsely (little image work, so I am rusty), and I still like what it stands for: a tool that grew in the commons, not a rented seat. Deepening that line — more polish, better defaults, less pain — would make me glad.

What sucks is starting by looking at Photoshop. That is the wrong side of the coin. It is like starting from a billionaire to learn what workers want. You will measure the wrong hungers. You will copy the palace and call it housing.

Open source, as I grew up with it, was about people: bring them together, federate them, decide the way forward from shared needs. Debian is not a clean-room Windows with Apple menus. It came up in a lot of clean air — different politics of distribution, different expectations of users, different idea of what a computer is for. GIMP is incomplete and awkward and ours. Improving GIMP is continuous with that weather.

Here the starting point was not a federation of common needs. It was an end product of a multinational — menus, shortcuts, panels already known — then a rewrite. How bad is that? Bad enough that I call it intellectual surrender, and still soft enough that I may be missing a bridge I have not walked. Cloning Photoshop’s muscle memory is continuous with Adobe’s product definition, even when the licence on the clone is MIT or Apache. Why put this wave of AI-assisted energy into 1:1 menu parity instead of deepening tools that never promised to be Photoshop? I do not have a neat answer. I have the discomfort of watching the word “clone” repeated as praise.

Mage-OS is not “the Magento you already know”

Closer to home: when Mage-OS answered Adobe’s Magento trajectory, the open-source core made the fork possible. Shared lineage is not the same as ceding the idea of the product. Mage-OS did not brand itself as “the Magento you already know how to use” as the moral of the story. The pride was in taking a different path — governance, independence, a refusal to treat Adobe Commerce Cloud as the only future — even while the codebase still spoke Magento.

Having code in common is not intellectual surrender. Handing over the definition of success — same menus, same shortcuts, same hands — is. That is the distinction I cannot shake.

Two outcomes I keep imagining — and a theory that chills me more

I am not predicting the future. Still, from where I sit, two endings feel more likely than “a better world arrives because the face matches”:

  • Adobe (or a neighbour) employs some of the talent that proved the map can be redrawn quickly; or
  • the first serious legal season — trade dress, contract terms, patents; clean-room talk does not erase “same face” — burns the experiment as a commons story.

There is a third line of doubts I keep hearing in research notes and in what many book writers (and songwriters) have been saying about generative models: that a clean room with Claude is not possible in principle. Classic clean room needs a wall — people who write the replacement must not have seen the original’s protected expression. An LLM sits awkwardly in that room. Nobody outside the lab can audit what the model ingested in training. Commentators on ArtCraft already ask the blunt question: what did the model read? If the answer is “we cannot know,” the alibi of the clean room wobbles. I am not a lawyer; this is the theory as I understand it, intended for further reading, not a verdict.

Here is the bitter twist, if that theory lands: authors and musicians spent years arguing that training on their work without consent was theft — a gun pointed at big tech. Are we now handing Adobe (and friends) that same gun, except aimed at open-source “suite clones” built with the same class of models? Fuck: basically, did we gift the multinationals the weapon book writers and songwriters forged against them? They must be laughing. Or filing notes. Either way, the commons does not look smarter for it.

Meeting them on their own ground is dangerous especially if you had better ideas. Hearing “Adobe-suite clones” on loop is, to me, already a little disturbing: it normalises the clone as the achievement.

I will be rude, because I have nothing to hide and this is not a party leaflet — it is a statement I like to make. Fairness talk that celebrates cloning Adobe’s suite feels, to me, like calling China a USA alternative: same imperial operating system, different branding — state capitalism with a red flag instead of an oligarchy with a blue one — and then pretending the rename is liberation. There could hardly be something more disturbing. If you rewrite the oppressor’s form and sell the rewrite as freedom, I do not see how something good is supposed to come out of that (here I am still on that provocative USA–China “alternative” — pointing at how some people read US posture in the Middle East and Latin America, for example).

On China more specifically — and on Tibet, silence, and ethics under big tech East and West — I wrote briefly in The Way of the White Clouds: ethics, silence, and the web we were promised.

I may be wrong

I am confused on purpose. Agency lock-in is ugly. Subscription traps are ugly. File interchange is work (for whom this is not clear enough: what I mean is that making PSD, DXF and similar files open and save correctly across programs — layers, quirks, round-trips — is real engineering labour, not a free side-effect of “looking familiar”). Familiar UX can be a bridge for people trapped in proprietary seats. A bridge that is the destination — “you already know,” forever — still looks like surrender to me.

Open source, for me, used to mean we believe in better ideas, not only better packaging of the old ones. If the craft’s scarce hours go into rewriting the suite you already know, I do not yet see how a better world is supposed to come out of that. I am writing it down so the doubts have a place to sit.

That sucks. In these moments I sadly feel only beer and rock & roll can save us (don’t get me wrong: I mean rock & roll in its anti-establishment posture — and if you think jazz, blues and punk are not part of what can save us, you had better read someone else’s articles).

Postilla — 28 October 2026

Someone told me I am nostalgic, and a fan of Linus Torvalds. I am not a fan of Torvalds. I value what he did. I have also heard too many accounts of how harsh he can be with others — and that sits badly with me. What reassures me is simpler: Linux will survive Linus. I have read open letters of his that reminded me, uncomfortably, of the colder side of other celebrated tech leaders. “Success” does not impress me when it comes with contempt for people. I value people who treat others as peers. That is what Debian does. Debian will survive Linus Torvalds. There is no way Linus Torvalds will survive Debian. No hard feelings — only the sense that a life spent always above the room, never as employee or peer, leaves a different mark.

Written with the help of an AI assistant.