Dan Catt (known online as revdancatt) is a print artist who draws with code. He writes generative programs, lets the algorithm spit out hundreds of SVGs and PNGs, then runs them through a Riso duplicator — an old machine somewhere between screen printing and mimeograph, where each colour gets its own drum, the design is "burnt" onto that drum, and sheets of paper are fed through one at a time. He doesn't post the work to social media; instead he keeps a public "dev diary" on his own site, logging it day by day — the afternoon he trimmed some business cards, the morning he noticed a drum was printing crooked. As of late April 2026, the diary is still being written, one entry at a time.

What's interesting isn't the clean code — it's where the machine misbehaves. Catt found that the yellow drum always prints slightly skewed: not a simple up-down-left-right misregistration, but the whole image twisted by a small angle. Stranger still, he suspects the amount of skew depends on how much ink is used — fine lines barely move, but lay down more yellow and it skews more. So he writes: "the next experiment is to use various amounts of coverage and see if the amount of skewing increases too." That sentence matters. He hasn't named the thing, filed it away, or solved it out of existence; he's letting it stay alive as an open question.

Then there's a bug he deliberately kept. While making his "80s Pop Roxy" pieces, an error crept into the code — but "it was a very appealing bug," so he took a snapshot, tucked it away, and only then fixed the original program. Months later he dug the "bugged" code back out, specifically to feed the Riso. It's an almost perfect specimen of a remainder: most people fix a bug and forget it, letting it settle into the construct of "resolved." Catt refuses to let it become construct, preserving and reusing it as a piece of leftover matter the system hasn't yet digested.

In SAE terms: the pen-plotter aesthetic — that whole "code, paper-feed, ink line" idiom — has by now been chiselled out, named, and absorbed by the art world. It is already-construct. It still has its own beauty, but that beauty is sediment left by a past chisel-construct cycle; it no longer pulses as a living remainder. Catt's Riso practice sits in the gap that hasn't been named yet. The algorithm asks for one clean vector image; the machine adds skew, misregistration, and ink bleed on top of it. Those additions are the remainder — never specified by the code, not yet theorised by the art world. Here the chisel is no longer the programmer's algorithm but the printer's own body, and what it cuts loose is exactly the part the machine computes on the algorithm's behalf.

Why does seeing it now matter more than seeing it later? Because right now it is still alive. Catt is measuring that skew publicly, day by day — and he hasn't found the answer yet ("not yet anyway, looks like I still have more experimenting to do"). The moment he writes the correction algorithm and polishes "generative Riso" into a repeatable style, a gallery will eventually absorb it and it becomes one more construct. What's worth watching is precisely this state — before it congeals into a name, while its logic is still growing. His stance of closing the shop, swapping selling for bartering, swapping social media for a dev diary, also keeps this remainder, for now, just outside the reach of the institution.

Han Qin (秦汉) · @hqinjarsy · SAE Judgment & Aesthetics