Restorations & Reconstructions

For the last seven years the lab has rebooted numerous work of born-digital media by documenting them via Grigar and Moulthrop’s Pathfinders methodology, which has culminated in a series of electronic open-source books entitled Rebooting Electronic Literature, and by conserving them through various restoration and reconstruction initiatives.

The former, documentation, involves no direct intervention into a work, but the latter, restoration and reconstruction, requires interventions to the code and other aspects of the work that may involve emulation, migration, and/or collection in varying degrees and, so, always results in new versions and, thus, editions of a work. We call the intervention into portions of the code (including changing linking structure) and/or aspects of the functionality of a work to make the work accessible again “restoration.” Such efforts entail low level media translation. [1] We call the complete rebuild of a work that affects its code and may also impact its functionality and presentation “reconstruction.” Such efforts entail mid to high-level media translation. Thus, digital preservation in these contexts involves unique conservation practices identified for the specific needs of a particular work. Both restoration and reconstruction reflect conservation activities aimed at keeping the work accessible to the public.

Below are the 10 conservation activities the lab has undertaken over the last three years.


  1. M. D. Coverley. “Fibonacci’s Daughter.” Self-published. 2000. Now at The NEXT: https://the-next.eliterature.org/works/1064/0/0/). The original sound files (MIDI) were converted/migrated into a contemporary program (mp3) and added to the work, thus replacing the outmoded format. The results were tested against the original work on legacy hardware and software. 
  2. All of the hundreds of Flash works restored via Ruffle. The process involved adding code to the server and the work in order to emulate the unsupported Flash Player. None of the original code or media was touched in the process.
  3. All of the hundreds of Flash works restored via Conifer. The process involved zipping the work into a WARC file so that they can be accessed on an emulated browser. No changes were made to the work’s code but in order for the emulation to be successful, the works’ source code needs to be handled by a different machine code, different interpreters and the container scripts. The URLs for the works were changed to reflect their presentation on Rhizome’s server


  1. Deena Larsen. “Kanji Kus.” Self-published. 1999-2002. “Kanji-Kus 2.0.” Electronic Literature Lab, 2019. http://dtc-wsuv.org/wp/ell/deena-larsens-kanji-kus/). This series of 17 interactive poems was originally created for the web in HTML with iFrames and image maps produced with Java Applets. The lab migrated and retained the images and text. The code was also migrated but rebuilt to remove the need for iFrames, and the image maps were programmed as JavaScript. At each step of the process, the reconstructed version was checked against the original work on legacy hardware and software.
  2. Annie Grosshans. “The World Is Not Done Yet.” Self-published, 2013. “The World Is Not Done Yet 2.0.” Electronic Literature Collection 4. 2022. https://collection.eliterature.org/4/theworldisnotdoneyet. This hypertext essay was created for the web with Adobe Muse. The lab migrated the images and text of the original but completely rebuilt the code in HTML, CSS, and JavaScript. The many 2D text animations that emphasize the elegiac tone of the work were recreated in JavaScript and CSS and checked against the original work on legacy hardware and software to ensure fidelity to the tempo of their appearance and movement. 
  3. Richard Holeton. Figurski at Findhorn on Acid, Version 3.3. Watertown, MA: Eastgate Systems, Inc., 2001. Figurski at Findhorn on Acid, Version 7.0. The NEXT. 2021. https://the-next.eliterature.org/works/1102/0/0/. A hypertext novel originally created with Storyspace and published on CD-ROM, the work’s code was rebuilt in HTML, CSS, and JavaScript. The lab migrated the text from the original but updated typefaces, edited text for typos, and created new images where needed. Because it is a classic hypertext from the Eastgate Systems, Inc., the lab built two modes of reading, one that updates the work for a contemporary audience and another that retains a semblance of the original Storyspace experience. The results were checked against the original work on legacy hardware and software, specifically the Apple iMac G3 (Bubble, circa 1998-2003)
  4. Thomas M. Disch. Amnesia. Redwood, CA, 1986. Amnesia 2.0. 2021. https://amnesia-restored.com/. This text adventure was created on the custom-built game authoring platform “King Edward” written in FORTH in a version called “Atila” for the Apple II (and later migrated to the Commodore 64 and the IBM PC). The preservation team rebuilt the work in HTML, CSS, and the text parser, TextEngine, a self-contained JavaScript library that the team heavily modified for the game. The game’s interactive map was produced with Leaflet, a JavaScript framework. Major changes were made to the interface, functionality, and presentation. For example, the physical media packaged with the 5.25-inch floppy disks were incorporated into the gameplay. And while the text from the original was migrated into the new version, the parts of the story omitted by EA but desired by the artist were added back into the story. It also offers four different visual experiences: Contemporary Mode, Apple II Mode, Commodore 64 Mode, and IBM PC Mode. The results were checked against the emulated version (DOSBox) and the original work on legacy hardware and software.
  5. Sarah Smith. King of Space. Watertown, MA: Eastgate Systems, Inc. 1991. King of Space Version 2.0. 2022. http://kingofspace.org/. This interactive game was originally created on the Hypergate authoring system developed by Mark Bernstein for Eastgate Systems, inc. and published on two 3.5-inch floppy disks. The preservation team migrated the text but consulted the two manuscripts and the author’s notes to add content for the characters’ backstories. Without access to the original code, the team rebuilt the work in HTML, CSS, and JavaScript. Major changes were also made to the interface, functionality, and presentation. The narrative was also streamlined to follow one major path of the story to provide a coherent gameplay. The results were checked against the original work on legacy hardware (PowerMac G3 all-in-one) and software. New images and sound were produced and changes were made to the mini-games to make them more intuitive and playable.
  6. Stuart Moulthrop. Victory Garden. Watertown, MA: Eastgate Systems, Inc. 1991. Victory Garden 2022 (Version 5.0). 2022. The NEXT:  https://victory-garden2022.com/. One of the classic early hypertext novels built on the Storyspace platform and published originally on 3.5-inch floppy disk and later on CD-ROM, Victory Garden 2022 was rebuilt HTML, CSS, and JavaScript from a modified Tinderbox migration. The text was migrated but the interface was completely updated. The original horizontal toolbar, for example, is re-envisoned as a vertical navigational map. A wholesale change to the hypertext linking structure results in Paths and Streams, with the former capturing the multilinearity of the original, and the former offering a episodic reading experience.
  7. David Kolb. “Caged Texts.” Unpublished. 1994. In progress. The philosophical hypertext essay was originally intended for inclusion in Kolb’s larger hypertext, Socrates in the Labyrinth, but never published. To reconstruct it, the Storyspace file was exported to Tinderbox and then to HTML by Bill Bly, author of the We Descend, who then tested it in different versions of Storyspace (vv 2.51b1 thru 3.90) & Tinderbox (vv 4.7.1 thru 9.1.0b542) to get the proper result. He also checked to ensure that its linking structure––the default and text links––was successfully migrated. It also required Bly to tag the guard fields and loops for building the navigation. Using this code and text, the lab is currently designing a contemporary interface for the work that is accessible across all computing devices.


[1] As Mariusz Pisarski and I argue, “the preservation of digital art and expressive writing, such as born-digital literature and net art, through the process of migration and emulation involves the translation potentially between formats, software, platforms, hardware, computer languages, and/or digital qualities in a way that impacts the human experience with such works. As we will show, it may or may not involve linguistic transformation, but always the underlying code is affected and may or may not result in changes to functionality and presentation. As translation of creative media, translation requires the interpretative intervention of the translator and, so, is an act of creation.” This is what we mean by media translation.

Dene Grigar is Director and Professor of the CMDC Program. She specializes in electronic literature, emerging technologies and cognition, and ephemera.