2003-08-27 Rolling notes Hypertext 03 http://www.ht03.org/ Programme http://www.ht03.org/programme.html All the papers are already online here: http://www.ht03.org/papers/ They're all easily readable. I'll highlight my favourites later. (We've all been given a book of them.) # PDFs of poster presentations are also all online: http://www.acm.org/sigweb/Ht03posters/ David Gibson's paper, "Community Reputation in Dynamic Hypertext" http://www.acm.org/sigweb/Ht03posters/Papers/gibson.pdf is also worth a read. It describes how to extract the community's opinion of a page, a community-specific pagerank. Imagine choosing your peergroup and when you're browsing having their aggregated opinion of each page you browse to given to you as a score. Possible! This thing can calculate in realtime. # there were some issues about scaling being mentioned on the stairs down from the [Ted Nelson] keynote. how's it going to be done with millions of documents and no URNs? but i think that's kind of missing the point. the way the universe solves this problem is by introducing locality and distance. you worry less about things at a great distance. zigzag creates its own space by introducing dimensions. if we stop thinking of nodes as documents, would there be a way to implement this in a vastly distributed way? i still have my concerns about transclusion though. it's so easy to retype small bits of text, and the programatic context of what i talk about as "a fish" is just as important as what's written down. transcluding must necessarily throw away information - often the context. --> most important bit of the keynote is that ordered lists are first-class in zigzag: this means that a *distance* is defined in this hypertext in a way it simply isn't in the semantic web vision. very important. how the world works. maybe it means things don't have to be so constructivist, because there are other properties/contexts rather than stated ones, and it's/they're n-dimensional: distance from everything else. that's emergent in the www, but at the very essence of zigzag. # Jeff Sonstein kept Hydra notes on "Mixed Reality Hypermedia" presentations... I don't know whether these are going to be online. The Dexter hypertext reference model is mentioned in there. # Dexter hypertext reference model http://portal.acm.org/citation.cfm?doid=175235.175237 "The Dexter hypertext reference model is an attempt to capture the abstractions found in a wide range of existing hypertext and hypermedia systems and those abstractions predicted to be included in future systems. Its development was motivated by the discussions at two small workshops held in the late 1980s. Those invited to these workshops represented many of the hypertext systems of that time. The first openly available description of the Dexter model appeared in 1990 [1]. This report was not widely distributed, but the Dexter model has become extraordinarily influential in hypertext research. This paper is a condensed version of the original description. A significant contribution of the original version of the paper was its inclusion of a formal specification of the model. The present version eliminates the specification, but adds informal descriptions of the model's functions and operations. Consequently, the original version of the paper remains an important resource for those specifically researching hypertext models, but the condensed version will be sufficient for readers with different interests. Every hypertext researcher should be familiar with the Dexter model, although it is clearly an open question whether the model can adequately represent the characteristics of tomorrow's hypertext systems. This paper does an excellent job of making the model available to a wide audience in an accessible and understandable form. It is required reading if you are contemplating hypertext research." # app to pursue: mspace by m c schraefel http://www.dgp.toronto.edu/~mc/ http://www.dgp.toronto.edu/~mc/projects.html http://wwwis.win.tue.nl/ah2003/proceedings/paper19.pdf is really interesting. a way of browsing classical music without having to know the axes we usually impose on it. prototyped now! more on mspace: mac os x application that lets people establish context by playing snippets of music and explore facets *without having to know the classification scheme*. you can pivot it on different dimensions and discover music. currently backs on to mysql and a set scheme of metadata, and a whole load of CDs from sony and universal free for academic purposes. you could swap that out and replace it with stuff the bbc owns the rights to. what a demo that would be, to mark up everything else! and how essential, given the creative archive project. need to tell people about this. (also need to talk to people about the topia project. imposing a narrative over search results. explain, not search.) # still thinking about zigzag. i really like the way the dimensions are so key to the thing. that anyone can build new dimensions. anyone can do link overlays. in a way, it's more fundamental than just a graph, that nodes are smeared out. # from a session i missed, and the Hydra notes run by Jeff Sonstein: === again i'm in the other room, but you may find www.aktors.org of interest for some real happening semantic web stuff. the manifesto is pretty compelling-and the real world work is also good stuff. Nick Gibbins who's here and from that group did a tutorial in this space yesterday. if you're new to the area, getting his notes might be a good thing. ===