FRESS: The File Retrieval and Editing SyStem

This page was written by Steven J. DeRose around 1998, and was last updated on 2003-03-28.

FRESS was the second system in the long and honorable history of Brown University hypermedia research. It was preceded by HES (the Hypertext Editing System). HES was one of the very first hypertext/hypermedia systems ever built, and the first on standard commercially available hardware (the IBM 360, with CMS). HES was built by Andy van Dam and his students, with a lot of input from visionary-in-residence Ted Nelson, who comments in Dream Machines (or is it in Computer Lib? I forget) that all the WYSIWYG word processors derive from HES -- but lost the hypertext!

Ted and Andy were undergraduate classmates at Swarthmore, along with Peter Schikele; which surely tells us a lot though I have no idea what. I still want to see the libretto of the rock opera they are said to have written, before such a notion as "rock opera" even existed (I've variously heard that the score was written by Peter Schickele or Dick Caplan). These three all seem to keep turning up in my life, though. In high school I was given a music award consisting of Schickele's the Definitive Biography of PDQ Bach; I ran into Ted in an elevator at ACM Hypertext '87, and have debated the interaction of markup and Xanadu versioning with him from time to time since. Andy introduced me to Lou Reynolds with whom I co-founded Electronic Book Technologies (later part of the late Inso Corporation). Inso in turn grew out of a sofware group doing spelling and grammar correctors, that I (and David Durand, my HyTime book co-author) had a bit to do with developing long ago. And just to complete this twisted picture, I was a Teaching Assistant for a while at the University of North Dakota, just down the road from Hoople and the U of SND at Hoople, where Schickele teaches his graduate seminar on "Originality Through Incompetence". Oy.

FRESS pioneered a number of features, such as unbounded document and node sizes; bidirectional linking; keyworded/attributed nodes and links; and, perhaps most important in the long run, the "Undo" command. You can finally read about FRESS, in an article by Andy van Dam and I in the MIT Press journal Markup Languages.

FRESS was not only a hypermedia system; it was also so good at text editing and formatting that it was used to typeset quite a few books, and had good enough IR and data structuring facilities that it was pressed into service for databases. NEH and Exxon funded much work with FRESS in educational hypermedia, and it was used to teach courses in poetry and other topics. For some time Ted disavowed FRESS. I suppose this is because it had no version management (and no revenue model). Real version management was far beyond the state of the art then; it arguably remains at the edge even now. On the other hand, FRESS did have the key features of bidirectional linking and (rudimentary) transclusion, which I don't think re-appeared until the mid-1980s (in Brown's Intermedia). In my experience, Andy has always been entirely respectful of Ted's vision and innovation, of his contributions to Brown's systems and to hyperetext in general, and of Ted's different views.

It is not widely known that FRESS still runs on the current Brown mainframe, though the hardware is vastly different from the machine on which I last assembled (!) FRESS in 1981 from nearly a million lines of source. David Durand built a Macintosh terminal emulator by reverse-engineering the data streams FRESS sends if you tell it you're on an Imlac graphics display station. We used his emulator to demo FRESS live at the ACM hypertext '89 conference, at WWW6, and at the 1999 "AndyFest" at Brown University. The demonstration data comes from a poetry course taught entirely with FRESS long ago, and includes annotations by Norm Meyrowitz's wife, a student at the time. Norm's name is familiar from his leadership of Brown's Institute for Research in Information and Scholarship, at Go, and at Macromedia.

The "Imlac" just mentioned was a high-end vector display on which FRESS shone: where it could do multi-window WYSIWYG editing of text and graphical hypermedia (these features were a bit harder to accomplish on literal typewriter terminals, which were much more common in those days). Perhaps its best feature was that it used a lightpen instead of a mouse. Pointing to text or menus on the screen and then stomping on the foot-pedal provided a certain catharsis unmatched by the modern mouse button. Four huge panes full of 9-point text did tax the vector engine's ability to draw little strokes, though, and every now and then I had to re-boot the machine by loading code using front-panel toggle switches.

I plan to add much more information on FRESS here later. In the meantime, some other references to FRESS are:

Oh, what the German speakers among you are thinking about the name is accurate.

If you ever used, or especially worked on, FRESS or HES, please drop me email, since I would like to gather more information about the history of hypermedia at Brown, and perhaps even assemble a collection of essays on the subject.

Back to home page of Steve DeRose or The Bible Technologies Group. or The Bible Technologies Group Working Groups. Or, contact me via email (fix the punctuation).