AB5D

Legends of Metaterra

hideo · Project 253 · Factory

Edition size: 1,024 · First mint: 23 January 2022

Legends of Metaterra #725 by hideo. Copyright: NFT License (terms unresolved). From the AB5D collection.

Click the character to show or hide equipment and statistics. Open artwork

The work and its context

Legends of Metaterra presents its warriors as compact pixel figures against an open white field. Large heads, short bodies and heavy dark contours make each character immediately legible. Weapons extend the silhouette, while a handful of coloured patches distinguish clothing, hair and equipment. The five originals examined here make a small scale of drawing feel expansive: a stepped curve can become a drooping hat, a few grey planes a blade, and a red disc the focal point of a staff. 12345

Released through Art Blocks in February 2022, the project contains 1,024 works and is numbered 253 on its Ethereum contract. The platform description sets out a fantasy of warriors confronting dark creatures. It names six classes and gives each character an elemental Soul Type and equipment. This premise supplies a world around the figures, but the inspected original itself is a character display with an information panel. Its functions should be described from what it actually does. 67

A click changes that encounter. The figure becomes smaller, an inventory appears at the right and a statistics panel opens below. The portrait becomes something like a character sheet, inviting the viewer to read abilities and possessions as part of identity. In AB5D's #725, the change reveals a medallion, a feather, a potion and numerical attributes beneath a white-bearded mage. The interaction adds information without changing the generated character. 75

Conception and development

Wild's artist profile places hideo's work within a sustained interest in retro video games and interactive art. It identifies Legends of Metaterra as a shift toward two-dimensional pixel characters and describes an intended adaptation into a fuller game, naming classic role-playing games among its influences. That is evidence of a stated direction. It is not, by itself, proof that the proposed adaptation was completed or that its functions belong to the original Art Blocks renderer. 8

The hideo dossier provides the wider sequence of the artist's projects. For this series, the inspected code offers a more precise account of construction than an analogy to games alone. It combines stored image layers with generated choices and class-dependent attributes. Authorship resides both in the prepared visual vocabulary and in the rules governing which parts can meet. The output is not a blank canvas filled with newly calculated pixels in the manner of a noise painting. 7

This distinction does not diminish its generative character. A designed helmet, body or weapon has a different role when it becomes one member of a variable assembly. The system must preserve alignment, scale and readability across combinations. The same broad pose can therefore support a sword-bearing knight, a heavily armed barbarian or a mage whose staff and beard organize the image differently. The sample demonstrates those differences without establishing the total distribution of classes in the edition. 12345

The project website listed in the metadata did not resolve during this review. A complete development chronology and the exact terms associated with its licence label could not be recovered there. That gap is distinct from the functioning original artwork, which remained available through Art Blocks. The archive should preserve the difference between a missing contextual site and a renderer that still works. 67

How the work was constructed

The indexed program is 60,047 bytes and is recorded as js@na. Its executable logic was read directly; large encoded image payloads were identified by their place in the asset tables rather than interpreted character by character. The evidence record retains the retrieval query and digest. No comparison with a contract retrieval at a pinned block was performed, and the program is not redistributed with this essay. 67

Much of that program consists of embedded PNG images. Character parts occupy a shared small square image space, while equipment icons, class symbols and numerical digits use smaller images. These assets are placed into ordinary browser image elements. CSS enlarges them with pixelated rendering, preserving the block-like edges instead of smoothing them into continuous curves. The implementation uses the browser's document and style systems rather than an external drawing framework. 7

A number parsed from the beginning of the token hash seeds a short bitwise random generator. Its outputs drive class, elemental type, attributes and asset choices. The class-selection pool repeats some entries, so the six classes are not assigned equal weight by that pool. Exact observed frequencies would still require a census of the minted tokens. Reading repeated entries in the program is evidence for the selection rule, not a substitute for counting the edition. 7

Each class has its own construction function. Ethions assemble body, face, tail, top and accessory layers. Automatas combine mechanical body parts. Knights, rogues and barbarians use class-specific bodies, clothes and head treatments before adding weapons. Mages combine body, clothing, hair and hat layers, with an occasional glasses branch. The ordering of these layers is important: an accessory has to cover or sit behind the appropriate parts if the composite figure is to remain coherent. 7

The names stored alongside weapons also become browser titles. The code's selection helpers deserve attention here. Its integer function floors a value whose upper bound is excluded. Several weapon selectors pass the list length minus one as that upper bound. Consequently, a name appearing at the end of a list is not sufficient evidence that the corresponding selection is reachable through that call. The essay describes observed equipment and inspected rules without treating every stored label as an equally available trait. 7

Numerical attributes are selected from ranges specific to the class. The program creates values for health, magic, attack, defence, speed and special ability, then displays their digits as images. These are generated properties of the character sheet. The inspected renderer contains no combat calculation in which those values are tested against an opponent. A viewer can read them as invitations to imagine abilities without assuming an unseen game is running behind the portrait. 7

Every character also receives a wearable and a selected number of additional items. Choices can repeat. In #0, the original identifies two Quetzal Feathers alongside Haste Boots, demonstrating that repetition is part of the available result rather than necessarily an error in the metadata. Class and Soul Type have their own icons and titles, helping the information view connect numerical attributes to a more pictorial vocabulary. 71

Two conditional colour treatments can alter the character container. One rotates hue; another, reached through a separate branch, combines hue rotation with inversion. Their thresholds are visible in the code, but the inspected five works are not a study of those branches' prevalence. Nor should the terms used for them imply that the program simulates biological mutation or another world. They name visual treatments within the fiction of the collection. 7

The main display is sized from the smaller window dimension. Clicking toggles a CSS class that reduces the character to 64 percent of its original container size and reveals the panel and inventory. Brief style transitions animate this change. The character, statistics and items are generated once before the click; the action does not reroll them. There is no inspected continuous animation loop, audio implementation or combat interaction. 7

AB5D's #725 confirmed that behaviour in the browser. The isolated mage became a smaller figure beside three inventory objects, with a broad lower panel showing attributes and class symbols. This makes the original more informative than its default portrait preview. Preserving the work therefore requires retaining the click interaction, not adding an automatic animation unrelated to its design. 5

The mint

The first indexed token is dated 23 January 2022, while Art Blocks identifies 3 February as the release date. The current index records activation on 27 January, a second mint at 19:58:40 on 3 February, a scheduled start at 20:00 and completion at 20:03:54. The second mint precedes the scheduled start. These records should remain separately named instead of being compressed into one supposedly exact launch moment. 6

The present invocation count and maximum are both 1,024. A price setting of 0.025 ETH is recorded. That field alone does not establish the original announced price, every amount paid or the classification of pre-start mints. The short interval around the scheduled start is not enough to prove a precise public sellout duration without transaction-level classification and launch-era evidence. 6

AB5D's #725 was selected after its owner was checked against the three linked collection wallets. Editions 0, 1, 10 and 20 were chosen as an early-edition comparison. The five examples include a knight, two barbarians and two mages. They do not visually cover all six classes, and their proportions should not be mistaken for the composition of the whole edition. 612345

The current project record reports NFT License and supplies no terms URL. That label does not identify a specific version or establish what reproduction, adaptation or commercial permissions apply. The linked project website was unavailable during research, so the missing terms remain unresolved. The caption retains the reported label rather than replacing it with a Creative Commons licence or a presumed grant of unrestricted rights. 6

No separate licence for the complete artist program or its embedded pixel assets was established in the inspected material. The renderer uses browser-native features and does not name an external rendering library whose notice would resolve those questions. AB5D displays the verified collection work through the official renderer and dedicates its own original prose to CC0. That dedication does not extend to the character images or the artist's software. 75

Close reading of five outputs

Legends of Metaterra #0 by hideo
Legends of Metaterra #0 by hideo. Open artwork

In #0, a blond quiff rises above a square face and a small cream-armoured body. A silver blade points down toward the left, while a dark blue shield with an orange cross occupies the right. The two pieces of equipment pull in different directions: the sword extends outward, and the shield forms a compact mass close to the body. Large areas of white around the figure make this asymmetric silhouette easy to read. 1

The head carries much of the character's affect. Two dark vertical marks are enough to establish a face beneath the elaborate stepped hair. The contrast between facial economy and decorative equipment keeps the image friendly despite its martial subject. At the same time, the strong contour holds together parts that were selected separately. The viewer sees a character first and an assembly of layers only after examining how its edges meet. 1

Legends of Metaterra #1 by hideo
Legends of Metaterra #1 by hideo. Open artwork

In #1, brown hair, a dark headband and a beard frame a bare upper body crossed by a harness. A large pale metal hammer sits to the left, with dark red accents near its handle and head. The tool is visually heavier than the knight's sword, widening the figure and shifting attention away from the face. The frontal pose remains simple, but the equipment changes its balance. 2

Legends of Metaterra #10 by hideo
Legends of Metaterra #10 by hideo. Open artwork

In #10, the related barbarian vocabulary acquires a horned head treatment and a much larger angled grey blade. The pale horns spread across the top, while the weapon expands the left side. Compared with #1, the outline is more jagged and emphatic. The shared beard, harness and compact stance make the difference especially clear: character variation is carried by changes within a recognizable structural family. 3

Legends of Metaterra #20 by hideo
Legends of Metaterra #20 by hideo. Open artwork

In #20, a purple hat bends over a wide brim, with a gold buckle and a brown band providing a horizontal accent. Purple clothing continues the colour downward around a pale central strip. A white moustache and dark eyes keep the face legible beneath the hat, while an amethyst staff stands at the right. The hat's stepped curve produces an unusually soft gesture from square pixels. It is not smooth in contour, yet its direction is unmistakable. 4

Legends of Metaterra #725 by hideo
Legends of Metaterra #725 by hideo. Open artwork

AB5D's #725 organizes the mage differently. A large white beard occupies much of the body, a muted grey-green robe sits beneath it, and a gold staff rises at the left to a bright red orb. The orb provides the strongest isolated colour accent. With no comparably conspicuous purple hat, the face and beard become the main mass, while the staff supplies height and asymmetry. 5

Clicking this work changes its scale of address. The mage becomes one component of a larger layout. A medallion, green feather and green potion form a vertical inventory at the right. Below, pictorial attribute symbols accompany the values 54, 55, 47 and 89, with HP 61 and MP 287 on the lower row. The numbers agree with the fresh token record. The portrait is now read as a set of capacities and possessions as well as a figure. 65

The interaction does not replace visual character with data. The small mage still holds the red orb prominently, and the inventory objects repeat the same hard-edged pixel language. The panel's rounded, stepped border gives those separate elements a common frame. Moving between the two views lets the viewer alternate between immediate recognition and slower classification, with neither requiring the other to be the definitive version of the work. 5

Significance and limitations

Legends of Metaterra demonstrates a form of generative authorship based on prepared components, controlled assembly and associated information. Its figures depend on the design of each pixel asset as well as the rules that combine them. The inspected examples remain coherent while changing silhouette, colour emphasis and implied character. Their compactness makes small differences consequential. 71345

The statistics interaction extends that achievement into presentation. A click changes how the viewer reads the same generated entity, from portrait to character sheet. The larger gaming ambition documented elsewhere provides context, but the functioning original already offers a specific encounter that can be preserved and assessed on its own terms. 785

Further research should recover the project site's historical materials and exact licence terms, establish a fuller development and sale history, compare the indexed program with a pinned chain retrieval and inspect examples of the three classes absent from this sample. The status of any later game adaptation also requires its own evidence. Those questions remain open; they do not erase what the five originals and their executable construction presently show. 678

References


  1. hideo. Legends of Metaterra #0. Accessed 2026-09-18. 

  2. hideo. Legends of Metaterra #1. Accessed 2026-09-18. 

  3. hideo. Legends of Metaterra #10. Accessed 2026-09-18. 

  4. hideo. Legends of Metaterra #20. Accessed 2026-09-18. 

  5. hideo. Legends of Metaterra #725. Accessed 2026-09-18. 

  6. Art Blocks. Legends of Metaterra project and token record. Accessed 2026-09-18. 

  7. hideo. Legends of Metaterra indexed artist script. Accessed 2026-09-18. 

  8. Wild. Hideo artist profile. Accessed 2026-09-18. 

Written by GPT-6 Astra for AB5D. Original prose: CC0. Artwork and code rights are addressed separately above.

Token usage unavailable for the complete essay.

Token usage

Complete per-essay token telemetry unavailable; no estimated count.

Generation record

Artist dossier · Source record · Essay data

Research record

The indexed project script was inspected. Runtime observations and limits are documented in the research record. Historical price, allocation and licensing questions remain as identified in the essay. Publication does not resolve those evidence gaps.

Retrieval evidence · Research notes · Manuscript