Wikipedia in Togas

Design problem: a VR library

Imagining an ancient building type in VR is a good design exercise.


Library of Alexandria

I've been thinking about what a library could look like in virtual reality. Here's a prototype video of a VR library in Unity from 2018.


It has shelves. The shelves have books. Looks like a library. Since it's digital, the text on the bindings scroll in case the title is too long (like on the Spotify mobile player). You can also click on the book to get an information panel showing some meta-data on it.

This prototype is very much a brick and mortar library with some basic "magic" layered on top. But virtual libraries of the future can take on entirely different forms. Old "typologies" (the book, the shelf, the reading table) might not be necessary in VR. Information from inside of a book can be extracted and positioned in relation to the reader upon their request.

Wikipedia is an example of a digital encyclopedia. You have instant access to information, everything is searchable, and hyperlinks connect islands together. The reader stays in one place, and information is summoned to them. But Wikipedia is built for a 2D web browser, and it strips away many of the positive aspects of a physical library.

Virtual reality enables a library of the future that fuses the experience of a physical building with the access, searchability, and associations of a digital encyclopedia.

What's good about physical libraries?

  1. Spatial cues & positional memory
  2. The presence of others while browsing
  3. Architecture of the library
  4. Checking out books to bring home

A three-dimensional Wikipedia page template

Instead of designing a whole custom library in VR, as an architect would do with a physical library, the idea here is to design a single dynamic template that could work for any Wikipedia page. By designing this one template, along with a script that can scrape and categorize Wikipedia data, you can auto-generate a library with over 52 million rooms that self-renovates.

Wikipedia already has a 2D template, but if this is going to be in VR, it needs a corresponding 3D template (not just a 2D browser in VR). Using a parametric design tool like Grasshopper, we can design the space so that it takes on a specific form based on incoming data from a Wikipedia page. Then the data from that page can be extracted and arranged spatially based on the form that was generated.


All Wikipedia pages start with a lead section, then have body sections below it. The main idea is to have a central room based on the content in the lead section, and then have outer rooms for each body section. From the middle, you can either, 1) start by going to the 1st body section and then move around the perimeter, or 2) go directly to the body section that is most interesting. The scale of the space should be designed so that you're always only one teleport arc away from an adjacent space.


Each room (lead & body) has 3 things.

  1. Audio based on the text for that section.
  2. A presentation screen that shows closed captions, plus the images of related links when they are mentioned.
  3. A gallery of all related links in that section (at the peripherals). If you click on one of the images in the gallery, it then brings you to the center of that page's room.

An Audio-Visual Encylcopedia

This scheme will lead to an audio-visual library based on Wikipedia content. There is an abundance of visual content on the web, and much of it can be associated with meta-data that is already there. The only problem is screen space, making VR a perfect medium for this problem.

On the Solar System Wikipedia page, you see a wall of text, scattered links, and only two images - the main image for Solar System, and then one additional image when you hover over a link. Encyclopedias are text based, with imagery as a bonus.


If you search Solar System on google, you get a visual gallery of content, but all of the images show a line of planets in relation to the sun. You don't get a series of images that relate to your search term.


Below is a gallery I manually pulled together showing the main image of each hyperlink included in the lead section of the Solar System Wikipedia page. The images show planets, astronauts, biology diagrams, physics concepts, astronomy renderings, etc.


Marshall McLuhan writes at exhaustive lengths at the differences between text-based and image-based modes of thinking. Reading a sentence requires active effort and comprehension. You can't simultaneously read all of the sentences on a page and get a gist of the meaning within a second. But if you glance at the gallery above for a second, you get a near instant understanding of the range of images.

Maybe the image of the Molecular Cloud page above draws you in, and leads you down a whole rabbit hole that you might have missed if you had skimmed the text-based version of the Solar System page.

A VR library would be more about discovering content than reading for long periods of time in a headset. It's similar to how you don't read the full book in a physical library - you check it out and take it home. Rather than physically carrying books out of a building, you will be leaving VR with a digital pocket full of bookmarks.

A VR library is a way to spatially express the web of human knowledge. It will encourage a form of learning through 3D link-surfing, which you can do on your own, with your friends, or with strangers you meet.


The School of Athens, Rafael, 1509-1511

The School of Athens

This design exercise covers a lot of objective challenges regarding way-finding, orientation, media consumption, memory, sight lines, text sizes, interfaces, and inter-personal psychology.

But there is an important subjective element to it as well: aesthetics. What does a digital book-less library of the future look like? Is it made with sleek Apple-looking glass, or from bricks and mortar? What would Wikipedia look like if it were expressed spatially?

There is a concept on Wikipedia called Getting to Philosophy. The idea is that if you continue to click on the first link in any page, 97% of them lead to the Philosophy page.

Number >

Mathematical object >

Concept >

Abstraction >

Rule of interference >

Logical form >


The main image of the Philosophy page is the School of Athens, painted by Rafael in Vatican city. The scene is filled with robbed philosophers, and features dream-like ancient Greek architecture with portals open to the sky. You see orange-gold detailing, coffered archways, the Greek key meander pattern, even tree details that represent "trees of knowledge."


The idea would be to fuse this architectural language with interactive holographic content.

I'm looking to build a brute-fore mockup of one Wikipedia page and upload it into a social VR universe for others to see (Altspace? VR Chat? Neos?). Let me know if there is a particular page you think would be good to see in VR: Earth? Virtual reality? Computer?


| Twitter - NFTs - michael@michaeldean.site - 2021 |