Where do new ideas come from? From conjunctions of needs, facts, experiences, and people. A personal example:

During late 1990 I was traveling out West, visiting computer research labs and contractors. A friend and colleague, Diane Q. Webb, was orchestrating the design of a hypertext-building and -browsing system. It was called "Hyperion" and ran on NeXT workstations. Creating it was a big, first-class software development effort costing many hundreds of thousands of dollars. I was a supportive user and part-time technical advisor to Diane, and so had become immersed in the challenges surrounding Hyperion.

(Aside: DQW was tiny, red-haired, quick, and driven, about a decade younger than me, highly articulate in her advocacy for new concepts. She started some good fires burning, but tread on countless bureaucratic toes (and land mines) in the process. Impatience and frustration led Diane to move on after a few years, first into the corporate software world, and then to form her own company, where she is today.)

In contrast to the Hyperion mega-project, for the California trip my wife had loaned me her wimpy notebook computer which she bought used for $100. With its 300-baud built-in modem I did a some email and USENET newsgroup browsing from my motel room. (Painfully slow, yes, but it did give one time to think in between reading messages.) At one point that evening a neuron fired and I realized that I had seen something Hyperion-like before in a completely different context: the "Info" mode of GNU-Emacs.

GNU was a disorganized effort to promote free software that grew up around Richard M. Stallman and his Emacs text editor. Emacs was an early windowing word processor. It had a built-in LISP interpreter that permitted anybody to write extensions and add new features — kind of like macros on steroids. One such extension was "Info" mode. Using "Info" anybody could take a technical manual and turn it into a set of pages, with cross-references magically transmogrified into jump-links. (N.B.: this was before the Web but after Bill Atkinson's Hypercard on the Apple Macintosh.)

All very well — but it had taken me several months to become semi-proficient at GNU-Emacs, and one could hardly expect a normal and busy person to make that kind of investment. Info files had to have all their tags and links edited in by hand, too laborious and risky a procedure to recommend in an office where work had to get done.

Aha! What we need is an easy authoring system for Info, says I to myself. That sounded do-able using Emacs extensions, but it also sounded much beyond my meager eLISP programming abilities. What to do? Ask somebody! So that night from my room I posted a note on the USENET "comp.emacs" discussion forum describing my semi-baked notion.

The message spread around the world over the next several hours. In Massachusetts the following day it caught the eye of Robert Chassell — LOGLAN/Lojban enthusiast, private pilot, Free Software Foundation officer, technical writer, eLISP hacker, and personal acquaintance of Richard Stallman. By coincidence, Bob had already done something vaguely similar to what I sketched out. Over a weekend he modified it and posted it on the Net. Yowzik! It was free, it was ugly, it was fun, and it worked. Bob named it "Para" mode, as a play on the Greek "hyper" prefix of hypertext. It expedited and quasi-automated the process of crafting Info files.

Para mode evolved a bit over months thereafter, and from 1991-1993 I made it my personal information management system, built networks of ideas, cross-indexed them, and wrote papers with that web of references as a foundation. It was a highly productive research environment for somebody with appropriate hardware, software, and background.

Alas, Para never caught on; there were at best a dozen users scattered about the known universe. We formed a little email discussion group and exchanged ideas for a few years, and then drifted on to other things. The Para mode learning curve was just too steep for non-GNU-Emacs-weenies to climb without ropes.

But the history of Para mode does highlight three deep sources of power that, in one form or another, lie behind all successful new ideas — and not just in the software realm. Those power sources are:

• a boundless extensibility — the potential to modify, customize, and reconfigure something far beyond the original conception. People's minds have that. So do human languages and mathematical systems. So does software ... provided the initial developers don't get selfish and lock the doors to change.
• a solid foundation — something to build upon that does at least part of the job well, something reliable yet elegant and æsthetic. It's impossible to do everything ourselves; we must rely on the work of our ancestors.
• a culture of sharing — so that several people, each with a piece of the jigsaw puzzle, can get together and come up with a solution. Sharing works well (or should) within a family, a small group, or a properly-run company. Ill-wrought "Intellectual Property Rights" laws, however, truncate the sharing process and give short-term profits to idea-squatters ... at the expense of long-term progress for all.

Three obvious factors, but perhaps not so obvious in their real-life implementation. Sic transit gloria Para.

A footnote: I still need a good hypertext authoring system. Surely one must exist by now — powerful, open, flexible, and productive. Any suggestions?

Tuesday, May 09, 2000 at 17:47:15 (EDT) = 2000-05-09

TopicProgramming - TopicPersonalHistory - TopicThinking - TopicProfiles

check out GZigZag @ http://sf.net/projects/gzigzag
it could be what you're looking for... it's a lGPL'ed implementation of Ted Nelson's ZigZag

for that matter, see one of my original ideas at http://purl.org/wiki/python/PeopleFilter as a possibly related (well, in the area of sharing) idea - Shae Erisson

Also check WikiIsIt on this server

(correlates: WikiIsIt, MalaproposDecisionmaking, NetworkNomenclature, ...)