AI: Summary
The meeting focused on academic document management in spatial computing environments. Participants discussed annotated bibliographies, efficient text organization methods, and visualization techniques. Frode Hegland demonstrated features in Reader and Author applications that support organizing text, creating annotated bibliographies, and managing references. The group also explored the concept of a “cube” interface in XR for organizing academic writing, with extensive discussion about how to represent references spatially and provide intuitive navigation. Different approaches to scaffolding spatial interfaces were considered, with emphasis on balancing freedom and structure. The meeting highlighted tensions between scrolling versus pagination, visualization techniques for documents, and methods to hide/reveal information efficiently. There was particular attention to how academic users might interact with citations and references in a 3D environment while maintaining workflow continuity.
Frode Hegland, Dene Grigar, Mark Anderson, Brandel Zachernuk, Adam Wern, Fabien Benetou, Karl Arthur Smink, Peter Wasilko, Rob Swigart, Tom Haymes
AI: Speaker Summary
Frode Hegland described his work with Reader and Author applications, presenting features for annotated bibliographies and linking to documents. He demonstrated text formatting options like fading less important text, which has been implemented in Author. Frode discussed his concept of a “cube” in XR for academic writing, where users can organize references around them. He emphasized the importance of “glanceability” and making information readily accessible rather than hidden by scrolling. Frode struggled with conceptualizing how to integrate writing and reference spaces in XR, acknowledging the complexity while seeking community input.
Dene Grigar provided valuable academic perspective on annotation practices, explaining how she summarizes sources and organizes them into folders based on relevance. She discussed three principles for object interactions in VR: maintaining natural affinities of objects, transforming objects from physical world, and breaking traditions for artistic purposes. Dene was excited about the possibility of three-dimensional hypertext in the cube interface, where links could traverse through the cube and maintain relationships when unfolded into 2D.
Mark Anderson discussed his reference management workflow using Bookends, explaining how he prioritizes fixing metadata before taking notes. He questioned how to handle transitions between viewing many references and focusing on individual ones in spatial environments, concerned about losing broader context. Mark distinguished between hypertext for thinking versus communicating, and suggested query-based approaches for organizing large collections of references. He noted that the “cube” concept needs clarity on whether it’s a mode control system or nested workspace.
Brandel Zachernuk contributed insights about pagination versus scrolling, noting he prefers paginated text on mobile for its “integrity” and clear context. He emphasized the importance of deconstructing print attributes thoughtfully rather than wholesale rejection, and highlighted how color coding (like in programming syntax highlighting) can convey semantic information effectively. Brandel suggested enhancing document visual cues beyond mere scrollbars to better represent document state.
Adam Wern made brief appearances in the meeting, mainly noted for his return which excited Frode. Adam was mentioned in relation to the left-hand menu design for the XR interface.
Fabien Benetou proposed using cube orientation to control scrolling speed in documents. He expressed concern about skeuomorphism when reproducing physical book affordances in digital spaces, wanting to balance familiar interactions with innovation. Fabien emphasized that scaffolding should be temporary, helping users adapt to new interfaces before fading away. He noted the cube’s technical simplicity and how work on it would benefit other primitive shapes in the interface.
Karl Arthur Smink shared his dissertation writing approach using LaTeX templates and BibTeX for reference management. He described using LaTeX comments as visual landmarks in lengthy documents and expressed interest in tools that could visually soften or hide text sections to improve navigation. Karl mentioned that he tolerates scrolling for short distances but finds it breaks down for longer documents.
Peter Wasilko described his research workflow using Devonthink, citation managers, and Tinderbox for mapping document relationships. He suggested using squarified treemaps for navigation and proposed a spatial interface with parallel walls showing concepts on one side and citation networks on the other, with interactive highlighting between them.
Rob Swigart briefly mentioned appreciating accidents and serendipity in research, suggesting crumpled notes could trigger random word associations similar to unexpected discoveries when browsing library stacks.
Tom Haymes cautioned about spending too much time on clerical aspects of bibliography management when AI could potentially transform how we interact with texts. He emphasized that user preferences should dictate content consumption rather than physical or technical limitations. Tom highlighted the importance of considering transitions between different “cubes” and collaborative possibilities, noting that XR’s advantage lies in transcending space and time limitations.
AI: Topics Discussed
WebXR was discussed primarily through Fabien and Frode’s exchanges about implementation possibilities. Fabien mentioned that the code for switching hand controls in WebXR would be technically feasible though somewhat challenging. The group discussed building their academic writing environment within WebXR, with concerns about collaboration features being difficult to implement directly. Tom expressed hope that improved collaboration features would eventually come to WebXR environments.
Gestures were discussed in several contexts. Frode referenced the Starfire Project from the 90s which used intuitive touch gestures where one finger would move an object while two fingers would create a copy (one finger staying behind). Fabian discussed interactions for the cube object, including grabbing, turning, and unfolding it. The interaction of pinching to spawn the writing cube from the dominant hand was mentioned. Dene suggested maintaining control over the cube by being able to turn it physically rather than having it follow the user automatically.
The Sloan work was briefly mentioned in the context of their current projects. Frode noted that Dene had helpfully narrowed down their academic focus to annotated bibliographies, connecting back to work Dene wrote in the Sloan project. Mark asked about setting up a test space with 10-20 objects as part of their Sloan work.
The Book was mentioned briefly when Frode discussed visualizing collections of papers like “The Future of Text,” which was described as a collection of papers similar to conference proceedings. This was mentioned in the context of how visualization approaches might differ for single papers versus collections.
Timelines were mentioned briefly by Mark Anderson as an example of a visualization that might be nested within the XR space without requiring the entire space to change.
VisionOS or Vision Pro development was mentioned briefly, with Frode noting that the Vision Pro already supports certain shared experiences, suggesting that collaborative XR might be possible even if not immediately implementable by their team.
Several other topics were discussed including:
Text formatting and presentation: Considerable discussion about scrolling versus pagination, with differing preferences among participants. They also discussed techniques like fading less important text to improve readability.
BibTeX and reference management: Discussion of different reference management tools and BibTeX field conventions, particularly around the “annotate” field.
Skeuomorphism in digital interfaces: Debate about reproducing physical book affordances in digital spaces, balancing familiarity with innovation.
Academic annotation practices: Different approaches to summarizing and organizing academic sources.
Spatial organization in XR: How to effectively arrange and interact with multiple references and documents in 3D space.
AI: Concepts Introduced
Scaffolded freedom: Introduced by Frode Hegland as a preferred term over “constrained” or “framed” for describing how to provide structure and guidance in XR interfaces without overly restricting users. Fabien expanded on this by noting that scaffolding should be temporary, helping users adapt to new interfaces before fading away.
Glanceability: Frode mentioned that he possibly coined this term in the 90s (edit: didn’t coin but likely first to use it on the Web for interaction design), describing the importance of having information within visual reach rather than hidden by scrolling or other mechanisms. He emphasized that information slightly below the fold is effectively hidden unless visible on screen.
Three principles for object interactions in VR: Introduced by Dene Grigar: 1) Maintaining natural affinities of objects (things should behave as expected in the real world), 2) Transforming objects from the physical world to do new things, and 3) Breaking traditions for artistic purposes when appropriate.
Three-dimensional hypertext: Dene Grigar described the concept of hyperlinks that traverse through 3D space rather than just across a 2D plane, specifically in the context of the cube interface where links could go through the cube to connect items on different faces.
Hypertext for thinking vs. hypertext for communicating: Referenced by Frode as a concept from Mark Anderson and Dave Millard’s paper “Seven Types of Hypertext,” distinguishing between hypertext structures optimized for exploration and idea development versus those designed for clear communication to others.
AI: People Mentioned
Bill Atkinson, mentioned by Frode Hegland and Brandel Zachernuk as a hero and reference for design discussions about scrolling, as they had lunch with him. Doug Engelbart, mentioned by Frode as having asked him to create text with color coding for semantic markup. Leibniz, mentioned by Frode as someone he was reading because Mike Witmore told him to. Dave Millard, mentioned by Frode as co-author with Mark Anderson of a paper called “Seven Types of Hypertext.” Bob Horn, mentioned by Frode as someone who worked with Brandel Zachernuk on a project that inspired some thinking about 2D spaces in XR. Andrew, mentioned by Dene Grigar as someone she’s been working with on three-dimensional hypertext concepts. Leon, mentioned by Frode as someone who might not be able to make the meeting.
AI: Product or Company Names Mentioned
Reader and Author, mentioned repeatedly by Frode Hegland as his applications for reading and writing. Bookends, mentioned by Mark Anderson as his preferred reference manager. BibTeX and biblatex, discussed by several participants including Mark Anderson, Karl Smink, and Frode as reference management formats. Zotero and Mendeley, mentioned by Karl Smink and Peter Wasilko as reference management tools popular among academics. Devonthink, mentioned by Peter Wasilko as part of his research workflow. Tinderbox, mentioned by Peter Wasilko for mapping document relationships. LaTeX, mentioned by Karl Smink and others as a document preparation system. Microsoft Word and Google Docs, mentioned by Karl Smink as having formatting issues he avoids by using LaTeX. ACM (Association for Computing Machinery), mentioned by Mark Anderson in relation to a paper he’s writing and citation mapping work. Overleaf, mentioned by Karl Smink as his LaTeX editor. LyX and Texifier, mentioned by Peter Wasilko as LaTeX-related tools. HyperCard, mentioned by Frode as a reference point for design. Kindle, mentioned by Frode and others when discussing digital reading experiences. Unreal Engine, mentioned by Dene Grigar as something they’re trying to use in their macOS lab but having difficulties with Apple compatibility. Objective-C and C++, mentioned by Karl Smink in relation to the compatibility issues between Apple’s development environment and Unreal Engine. Obsidian, mentioned by Karl Smink in a chat message about their commercial license becoming optional. Vision Pro, mentioned by Frode as supporting certain shared experiences. Starfire Project, mentioned by Frode as an example of intuitive gesture design from the 90s.
AI: Agreements & Disagreements
Agreements:
- Multiple participants agreed on the need for better organization of references and documents in spatial environments
- Mark and Frode agreed on the importance of distinguishing between hypertext for thinking versus hypertext for communicating
- Several participants agreed that the cube concept is a good starting point but needs further development
- There was general agreement that academic writing in XR needs to balance free spatial organization with structured presentation
- Dene and Fabien agreed on the technical feasibility of implementing 3D hypertext through the cube that maintains relationships when unfolded
- Many participants agreed on the importance of having multiple ways to view and organize information
Disagreements:
- There were differing preferences regarding scrolling versus pagination, with Mark expressing a preference for scrolling while Frode and Brandel noted concerns about hiding information
- Some tension between reproducing familiar physical affordances (like book pages) versus creating new digital paradigms
- Different approaches to reference management were described, showing disagreement about optimal workflows
- Some disagreement about how much structure versus freedom the spatial interface should provide
- Mark raised concerns about losing context when focusing on individual references, while others were more focused on the presentation of individual items
AI: Other
The meeting revealed an interesting blend of practical application development and theoretical exploration. The participants are actively building tools for academic writing while simultaneously philosophizing about the nature of text and reading in digital environments. There’s a palpable tension between honoring traditional academic practices while pushing into entirely new territory with spatial computing.
The role of Dene Grigar as an academic practitioner appears particularly valuable, as she grounds some of the more experimental ideas in actual scholarly practice. Meanwhile, developers like Fabien and Adam provide technical reality checks about implementation feasibility.
It’s notable how the group maintains enthusiasm despite acknowledging the enormity of the challenges they face. Frode’s comment about “literally pulling my hair out” before these meetings, followed by expressing gratitude for the discussions, highlights how the collaborative process helps navigate complex design problems.
The meeting’s format – allowing for wide-ranging discussion that occasionally goes off-topic but returns to core questions – seems to generate valuable insights that might not emerge in a more structured approach. This mirrors their approach to spatial computing itself: providing scaffolding while allowing for creative freedom.
Chat Log URLs
https://obsidian.md/blog/free-for-work/
https://ctan.org/pkg/knowledge
https://vanwijk.win.tue.nl/stm.pdf
https://a.co/d/iCO6auF
https://www.lyx.org
The chat log shows active parallel discussion alongside the verbal conversation. Participants shared links to relevant tools and resources, with notable mentions of Obsidian making their commercial license optional, the Knowledge LaTeX package, squarified treemaps for navigation, a book on interaction design lessons from science fiction, and LyX as a LaTeX GUI.
The chat featured real-time reactions to spoken points, with participants expressing agreement through emoji reactions. Several side conversations developed, particularly around reference management tools (Zotero, Mendeley) and LaTeX workflows. Karl Arthur Smink was particularly active in the chat, sharing his perspective on scrolling and his LaTeX workflow. There was notable discussion about the transitional nature of text management, with Tom Haymes arguing change is inevitable while Karl questioned whether change is always necessary or beneficial. The chat also captured quotes from discussions about print versus digital, with Mark Anderson clarifying he likes print but is frustrated by unhelpful echoing of print conventions in digital spaces. Overall, the chat facilitated a richer conversation by allowing participants to share links, respond to ideas, and develop parallel threads without interrupting the main discussion.