Archive of UserLand's first discussion group, started October 5, 1998.
Top Down, or Bottom Up?
Author: Paul Snively Posted: 9/17/1999; 8:48:59 PM Topic: Today's scriptingNews Outline Msg #: 11250 (In response to 11225) Prev/Next: 11249 / 11251
Dave Winer wrote:I just got off the phone with David Strom who is on the Invisible Worlds advisory board. The take-away message is that they're just getting started, and probably haven't thought very much about making their XMLizations public, but I agree it is the first step, it's what I want to see first, and then let the explanations and visions proceed out from there.
Certainly tossing out the DTDs and letting us technical types play with them is one approach to things, and as far as encouraging third-party development is concerned, it may prove critical in the long run.
But another equally valid approach is to say "all the structured data in the world isn't very useful without at least *some* applications to manipulate it," and to not expose much, if anything, until *some* applications exist. There may or may not also be a valid argument (or, more pertinently, a valid business model) that says "we don't just do the data structuring; we provide a middle layer that makes manipulating that data programmatically as simple as possible, and we hope developers will use our API to do that." Real-world customers, investors, and reporters will *still* want to see some apps, though, otherwise they won't be able to get it.
And in the final analysis, you have to appeal to your investors, whether VCs or angels or whomever, *very* few of whom would be qualified to judge an idea like Invisible Worlds from, e.g. the XMLization of EDGAR--and besides, the XMLization of EDGAR isn't the Big Idea anyway; it's just a conveniently complex testbed.
I like where Invisible Worlds is going: I suggested something very similar to it (conceptually) to a local enterpreneur over a year ago, but we lacked the resources to make a go of it and I couldn't figure out how to explain the concept to a non-technical audience... which convinced me that I wasn't the guy to pursue it.
So I'm happy that someone is finally seriously tackling the metadata taxonomy and interface problem.
Paul Snively
<mailto:psnively@earthlink.net>
This page was archived on 6/13/2001; 4:52:43 PM.
© Copyright 1998-2001 UserLand Software, Inc.