August 12, 2004

HT04 Conference Notes: Day 2

by Noah Wardrip-Fruin · , 5:48 pm

I presented this morning, and so couldn’t blog — but here’s a start at the afternoon. There will be more filling in, of both pages, I hope, before I fly to Helsinki tomorrow.

“Saving Private Hypertext” by Marshall and Golovchinsky. (Obviously, related to Acid-Free Bits and the ELO PAD project.) Starts with an image of the contents of Uncle Buddy’s — she can’t read the floppies, or even play the audio tapes, so all she can do is read the documentation. Then she saw that the web version of Forward Anywhere had a problem. She went to look at it, had a hard time reading her own code, and then finally discovered that the directory was protected by a password she’d forgotten.

She tells us about the perspectives of “digital optimism” and “radical ephemeralism” (which are both problematic). She got some pretty skeptical reviewer comments, and shares a good one — “you aren’t suggesting that, come 2054, no one will remember how JPEG works, are you?” She’s factoring out the issue of storage media. (She points out in passing, however, that most of us don’t store our CDs appropriately, for example.) She notes the RLG report that outlines three approaches to digital preservation: refresh, emulate, and migrate. (She notes that generations of emulation have not preserved Elvis!) Then the specter of copyright raises its head, but also gets set aside.

But, given all this even, literary hypertext presents even more challenges. She puts up a good list, ranging from preserving machine time to enabling/precluding originally possible/impossible readings. Some strategies get outlined, ranging from reimplementation (full preservation) to saving component parts (partial preservation). She gives a list of component parts, some of which are done (metadata, e.g., MARC) and others are “can of worms” (behavior, e.g., fluid hypertext). She mentions the idea of preserving the reading experience (e.g., with video) and going further (e.g., capturing multiple readings using hypervideo). And then there’s the question of who is going to do the work: readers, authors, publishers, libraries, philanthropic concerns (she mentions the ELO), and offshore service providers (everything expensive can be done offshore for cheap, using brute force, right?). Obviously, AFB is aimed at the authors part of this — and this is something that they’ve thought about as well. Authors can help, from their POV, by: using virtual machines to implement behavior, using external link representations, use existing means to record the work (and more I couldn’t catch).

Next paper is Andruid Kerne presenting a multi-author paper on “compositional hypermedia.” In their system their letting people create compositions, and one of the issues is the representation of surrogates — they think the right approach is to use sampling, not thumbnails or (just) summaries. The program is mixed initiative, with the human taking actions (to choose and arrange content) and the agent taking actions (to do the same things). The elements all connect back to their original context. This seems in some ways similar to work in CollageMachine, but more visually comingled (more blending) and with more generative capabilities (more the agent can do, and does interestingly). There’s a real focus on the history of interaction, and part of what’s new in this is that time can run both directions and there’s the ability to record parts of it, and a job-shuttle history slider like VKB’s (but with preview, keyframes, caching architecture). You can go back and change history, so that something that was erased/occluded from earlier is now preserved, brought forward into the current version. Check it out at

The next paper, presented by Hugh Davis, was a case study of a professor who decided to reimplement Microcosm (in which he had authored many educational materials) rather than port the materials to the Web. Hugh had hoped to find that this was done because the professor was excited by all the cool things that Microcosm did that the web doesn’t yet. But instead he found that it was the authoring facilities of Microcosm that had commanded the man’s loyalty. So, rather than a case for digital media preservation, he found himself making a case for better authoring tools for large web projects!

And now I’m presenting again — on “Scholarly Hypertext: The HT’04 Experiment and Beyond” — so blogging goes on hiatus again for a while.

One Response to “HT04 Conference Notes: Day 2”

  1. vika Says:

    Selfishly, I hope you blog your own talks eventually: scholarly hypertext is of particular interest to me. ;)

Powered by WordPress