Maps

The work and its context

Maps makes a field of circles behave like an unstable arrangement of solids and openings. Some forms seem filled, others hollow; a row of dots can become a perforation in a coloured surface or a string laid across it. The repeated verticals suggest an underlying order, but circular interruptions keep changing how that order is perceived. In the five originals examined here, the same broad vocabulary supports both fine, busy pattern and blunt areas of colour. ^{1 2 3 4 5}

John Provencher's project is numbered 241 on its Art Blocks contract. Its current edition contains 90 minted works. The title does not identify a surveyed location: these are invented fields whose environmental names belong to the colour system. A work called metallic forest need not depict trees or metal. Its greens and greys offer associations through which to read a pattern, while the pattern remains available as an arrangement of shapes. ^{6 7 5}

The visible token hash at the bottom is unusually consequential. It places a long identifying string beside an image derived from that input, making the relation between identifier and picture part of the presentation. The hash does not explain the image to the eye, but it keeps the work's particular origin in view. Even an apparently decorative sheet has a specific computational identity. ^{7 1 5}

Conception and development

The artist's project site connects Maps to graphic musical notation and sprite sheets. A score can propose several realisations, while a sprite sheet arranges states of movement without itself moving. The same site relates dithering to both digital bitmaps and physical halftone printing, and links the colours to Pantone references. Its catalogue includes two-colour screen prints of selected outputs, described as eighteen by twenty-four inches on white paper. These are documented print offerings, not objects physically examined for this essay. ⁸

Those references help explain why repetition here does not feel like a simple tiled ornament. The circles can be read as related states rather than identical units. A sparse circle beside a congested one invites the eye to imagine a transition between them, even though no animation is taking place. This is a visual interpretation supported by the originals, rather than a claim that every column encodes a particular movement or musical instruction. ^{1 4}

The print connection also makes the use of SVG more than a neutral delivery choice. The program converts an intermediate raster field into discrete vector elements. Large masses, holes and dots can therefore be understood as separate graphic marks rather than only as a compressed screenshot. Their sharpness exposes the character of the construction: a curved boundary can be visibly built from small repeated pieces. What might look like a smooth circle at a distance becomes scalloped on closer inspection. ^{7 3 5}

How the work was constructed

The full indexed script is ordinary browser JavaScript, listed as js@na. It uses native Canvas and SVG facilities rather than a named drawing library such as p5.js. It begins on window load, constructs its intermediate image, and schedules a later processing stage. There is no repeating animation loop, sound system or interaction handler in the inspected program. The original pages examined here resolve into still compositions. ^{6 7}

Random decisions come from a class that splits the token hash into two pseudorandom streams. It warms those streams and alternates between them to produce bounded choices. These choices establish grid dimensions, gradients, sampling intervals, colour pairs and several drawing modes. A given decision is made within a restricted range; the apparent profusion of little marks does not require a new arbitrary choice for every visible dot. Much of that detail emerges from processing a simpler field. ⁷

That field is created internally on a 180-by-240 canvas. The program does not import a photograph or a geographic map and reduce it to dots. It lays out circular regions in a grid, using either one region or arrays whose height exceeds their width by one: one by two, two by three, or three by four. Each circle is filled with a radial gradient whose centre can shift relative to the circle. Moving the gradient changes the distribution of light values while leaving the broad arrangement recognisable. ⁷

The distinction between circle and gradient centre is important. A circular boundary can contain an off-centre concentration of brightness. When that field is later sampled, one part may yield dense marks and another broad apparent emptiness. The final image's asymmetry can therefore arise inside a regular scaffold rather than from abandoning the scaffold. This is why identifying a grid is only the start of understanding the composition. ^{7 1 2}

A modified error-distribution pass processes the grey values. It uses the familiar fractional neighbourhood of Floyd–Steinberg dithering, passing adjusted values to the right and into the next row, with some downward contributions conditional on selected parameters. Its error calculation includes an additional divisor, and the routine does not simply replace every current pixel with a textbook black-or-white result. Calling it modified is substantive: the resulting values become material for later graphic decisions. ⁷

The program then reads ranges of those values through two different drawing routines. One emits coloured circles, implemented as ellipses with equal radii. The other emits coloured rectangular cells that can contain circular cut-outs made with SVG masks. Some cells have no effective hole. Selected brightness bands determine which kind and size of mark appears. The picture thus depends on a sequence of translations: gradient to processed values, values to categories, categories to geometric marks. ⁷

The two SVG layers can be sampled at different intervals. A larger interval changes the scale and spacing of the visible primitives. This helps explain why one output can look finely perforated while another presents large bead-like chains and blunt blocks. The order in which the layers are appended also changes in a branch of the program. A simple description of two colours printed in one fixed order would miss that compositional choice. ^{7 3}

Five base environment names and five accent names supply the usual colour pairing. They include forest, ocean, earth, void and nether on one side, and scorching, green, floral, dune and metallic on the other. The colours are specified in HSL, with additional special branches. These labels organise a space of colour relationships; they are not evidence that the program models an ecosystem. No geographical measurements, weather data or ecological simulation were found in the inspected procedure. ⁷

The portrait output is fitted to the window with padding and a pale or specially selected ground. The two SVG elements and the hash label are assembled into the visible container. Hidden canvases and an image element serve the intermediate processing. A short timer separates stages, so preservation involves browser behaviour as well as the retained vector appearance. The timer is a staging mechanism, not evidence of animation. A pinned comparison with the on-chain script and a systematic cross-browser rendering test remain future work. ⁷

The mint

The index records the first token on 15 January 2022 at 09:49:54 with a zero offset. Activation follows on 17 January. The scheduled opening is 21 January at 20:00, while the second token is recorded at 19:59:19 that day, 41 seconds earlier. These are distinct fields describing different events. The first token should not be presented as proof that ordinary public minting began six days before the scheduled opening. ⁶

The current completion field is 7 May 2022 at 03:09:04, well after the January schedule. The current invocation count and maximum both read 90. The artist's site also gives a collection size of 90, but that agreement does not by itself establish the original announced maximum or rule out changes during the mint. A historical account should preserve that distinction rather than infer an unchanged sale configuration from today's record. ^{6 8}

The observed price setting is 0.18 ETH. It is not independently verified here as the opening price or the amount every recipient paid. Separating artist, administrative and public mints would require transaction-level classification, and no exact public sellout duration is asserted. The available chronology nevertheless rules out describing the recorded completion as an immediate consequence of the scheduled opening. ⁶

Copyright and licensing

The project's current artwork label is CC BY-NC 4.0. Its record supplies no separate terms link, so the named Creative Commons deed is linked directly. The deed permits sharing and adaptation subject to attribution and the noncommercial condition, including indication of changes. A complete history of earlier artwork terms has not been reconstructed. ^{6 9}

No separate comprehensive reuse grant for the artist's script was established in the inspected material. The script is analysed here rather than republished. No external drawing-library dependency was identified in this program; its Canvas and SVG calls should not be assigned a p5.js licence merely because other Art Blocks projects use that library. The artwork designation, any permissions for artist code and AB5D's CC0 dedication of this original prose remain separate matters. ⁷

Close reading of five outputs

The sample comprises early editions #0, #1, #10 and #20, plus AB5D's #7. Their complete identifiers are 241000000, 241000001, 241000010, 241000020 and 241000007 on the project's Ethereum contract. The original generator pages were opened and inspected as finished stills. The treasury example was verified against all three linked AB5D wallets before visual selection. This small sample supports comparison, not a statistical account of the edition. ^{1 2 3 4 5}

In #7, green vertical bars run through a pale field occupied by two large grey circular forms. The upper form is divided by narrow green and white interruptions and bead-like perforations. The lower one is much closer to a solid grey disc, disturbed by a slender pale vertical break to the right of its centre. Their shared scale makes their different degrees of continuity especially apparent. One seems to be coming apart through a curtain of bars; the other seems to hold that curtain back. ⁵

Looking more closely, neither circumference is a smooth independent outline. Small repeated round marks build the edges, and grey beads also punctuate the top of the field. This correspondence joins the large forms to the fine structure surrounding them. The grey can read first as a flat opaque mass, then as an accumulation of the same small elements that elsewhere remain separate. That shift gives the image depth without requiring conventional perspective or a depicted object. ⁵

Edition #0 is much busier. A strong dark blue field carries orange fragments and a multitude of pale dots. Twelve circular regions can be followed through the three-by-four arrangement, but they are not drawn as twelve equivalent medallions. Some are nearly empty blue areas, others congested with columns of dots, pale bars and orange accents. Their boundaries become legible through changes of activity. A blue circle may seem cut out of the pattern rather than placed on top of it. ¹

The pale vertical chains keep the whole sheet ordered even where individual circles become difficult to separate from their surroundings. Orange works as intermittent punctuation: short steps gather at edges, then disappear. This makes the eye travel between broad recognitions and local discoveries. The grid provides enough regularity to sustain that movement, but the uneven distribution of marks prevents a single scanning rhythm from explaining the entire sheet. ¹

In #1 a much larger rounded region occupies the centre. Olive replaces blue, while pale vertical channels and dot fields surround orange strips. The large curve is established chiefly by a change in texture rather than a clean enclosing line. Its right-hand area becomes densely perforated; the left curve has quieter olive passages. Compared with #0, the image feels more monumental because there is one major form to negotiate, even though its surface remains divided into many small events. ²

Edition #10 makes a different use of scale. Bright green bead chains and broad green masses cover a pale field, with olive angular blocks visible beneath or between them. The repeated circles are still suggested, but coarse marks merge into emphatic silhouettes. At lower right, conspicuous round white openings interrupt olive cells. Here the primitive geometry becomes the main subject: the viewer can readily see how a round hole and a rectangular support exchange positive and negative roles. ³

The forest-green #20 is a useful counterpart to #0. It also presents a three-by-four arrangement, fine pale perforations and orange accents. Empty green circular regions alternate with densely worked ones, and vertical dot chains connect the intervals. The similarity invites close comparison, but it should not be called a simple recolouring: the indexed pattern labels differ, and the arrangement of marks is not identical. Green changes the associations available to the viewer while the shared construction makes those differences easier to locate. ^{4 1}

Significance and limitations

Maps is strongest where its economical ingredients refuse one stable reading. A circle can be a mass or an absence; a vertical string can organise the field or cut through it; a small dot can remain distinct or join its neighbours into a larger body. These shifts make the work more than a demonstration of a dithering technique. The technique supplies conditions for a perceptual exchange between the part and the whole. ^{1 2 3 4 5}

The five inspected outputs also show why the intermediate field matters. Their differences are not exhausted by named colours or counts of circles. Sampling scale, local grey values and the ordering of geometric layers change the balance between intricate surface and broad silhouette. Further work could examine printed examples, recover launch-era configurations and test rendering across browser environments. The present account establishes a narrower but substantial achievement: a compact program turns a simple graded scaffold into pictures whose apparent structures remain open to repeated looking. ⁷