Go to content Go to navigation Go to search

The OTHER reason(s) OPML is succeeding and rocks the Net · Nov 28, 01:09 AM by David Mercer

This is of course on top of it being open and usable by both mortal humans (non-programmers!) and machines (software, written by programmers). So here’s a little story about it….

Back in the days when Gopher ruled supreme, and the Web was just a little Cern linemode client icky-ness, where you ‘clicked a link’ by typing the number by it at the bottom of the screen, Gopher had structure.

You could search gopher-space, and someone had categorized that information into a hierarchy just by dint of putting it up on a gopher site: gopher sites were merely hierarchical menus that had content as the leaf nodes.

If I’m ever famous from programming, my personal “Bill Gates 640k of Ram/Ken Olsen the PC market is 5 computers” moment will be when I said, because of the chaotic nature of the Web and it’s shitty interface making it impossible to find anything compard to Gopher that “gopher rules, the web will never amount to anything” or roughly thereabouts (circa 1993…I’d had a couple-few years to develop the opinion by then).

Veronica, the search engine of the already structured gopher space, is how one found ANYTHING pretty quickly. Only google has finally caused the web to reach or exceed the relevancy of hits for web search, IMHO, and now even it more and more gives me searchs that are more and more polluted by marketers and fake portal and republication sites.

Doesn’t gopher sound a LOT like opml? Opml makes little user-created islands of hierarchical order out of the chaos of the Web, at least for those cases where such a structural ordering makes sense. That’s why yahoo’s taxonomy ruled supreme in the early days, the intentional design of the information space’s hierarchy to fit the subject by a human.

Opml let’s this be done by individuals, where ever it makes sense to, and rss let’s us move time ordered data around, including versioning information about opml hierarchies, either through an updated entry in an rss feed containing the link to the new version or indeed wholesale transport of updated versions via sending an entire opml document down an rss feed!!

Note that proprietary and open protocols and formats have been made to do all of these things, including quite a few others also based on xml, but that rss and opml were the first to align all Open-ness, non-programmer usability, and xml all at one go.

Kudos to Dave Winer, and to naysayers: transport your ‘perfect’ format document or protocol in leaf urls or on top of rss/opml and quit yer bitchin’!!

Structures Back!

-David Mercer
Tucson, AZ

References: http://www.scripting.com/2005/11/25.html#When:11:06:37PM
http://eirepreneur.blogs.com/eirepreneur/2005/11/why_opml_is_win.html

UPDATE: I must politely disagree with my commenters in a few respects.

1) How many non-geeks have ever generated a valid XHTML document by hand? How many web designers have?

2) Well the chaos of the web is good, don’t get me wrong, and yes brittle centrally managed hierarchies are bad from a pure-users standpoint. OPML get’s the best of both the web’s chaos and info hierarchy at the same time.

3) see, again, NORMAL users can generate them by hand, and it’s trivial to import/export it. And it’s still xml, it’s just ghetto xml….you sayin’ sumthin’ to my DTD? Huh? :-)

4) It’s already widely deployed in various apps, so developers can use it more easily….and yes, as it’s trivial even ‘bad’ programmers, the quick and dirty, just get it done and forget theoretical cleanliness programmers, can hack it onto anything because of that.

5) The confluence of events and timing just happened to break his way, and rss and opml rode the blog-wave up the ecosystem.

6) Yeah, it’s so simple anyone else could have done it, but they didn’t, did they? they made some uber-format that could express all of the freakin’ internal state of Their New XML Enabled Love Child, oops I mean program/web app.

7) Keep the tone civil, you’re in my house…and I was too fucking poor and living in a backpack to go to those wonderful conferences linked below, and being un-degreed I starved a LOT in those years, wandering in the wilderness (LITERALLY in the early 90’s a bit during Homelessness Bout #1…only metaphorially in the 80’s) telling everyone “someday, kids will email their grandparents and they’ll love it cause kids hate to write and they actually WILL this way!” and being looked at like I’m insane….and seeing That Exact Thing happen over and over again years later doing tech support.

Normal Users Can’t Write Other xml Formats By Hand…not a ONE of them, I dare you to produce in person for me a user who can’t program who can generate validating xml-ish formats other than opml or rss. Bring one to my porch and I’ll give you a Free Cat from my neighbors mess of them. Really. I dare you. I don’t think ANYONE could find one.

UPDATE2: Imagine that each opml file is a detached gopherspace node floating in the Web….that’s the best way I can find in words to describe the picture in my head! :-)

  1. Congrats on the Dave Winer link!!


    richard    Nov 27, 11:32 PM    #
  2. Much of the utility of the Web comes from the fact that it is a web, not a tree. We should be looking to taking advantage of that with our data, not rolling back to unsuccessful approaches. (X)HTML can already do everything OPML promises. Is OPML really more usable than HTML?

    Having said that idea of Gopher implemented on top of XML and HTTP is an interesting one, but as I noted a little while ago there have been advances over the past few years, see Gopher NG:

    http://dannyayers.com/archives/2005/07/14/gopher-ng/


    Danny    Nov 28, 12:06 AM    #
  3. Please take care to read the 1992 GopherCon trip report from Prentiss Riddle to see what the gopher vs. www discussion was like back then.

    http://iubio.bio.indiana.edu/soft/util/gopher/gophercon1.txt

    Ed


    Edward Vielmetti    Nov 28, 12:19 AM    #
  4. Truly wonderful Ed, thanks.


    Danny    Nov 28, 07:01 AM    #
  5. loved this piece, and it stimulated a further rumination, at http://blogs.law.harvard.edu/jim/2005/11/28#a1072

    thanks, jim moore


    jim moore    Nov 28, 02:12 PM    #
  6. “Imagine that each opml file is a detached gopherspace node floating in the Web….that’s the best way I can find in words to describe the picture in my head! :-)”

    Hold that thought. I think you’re picturing the idea of data on the Web, generalised from just documents to stuff we can use computers on, however we like. Interwoven information structures as simple or as complex as we want.

    Now imagine how good it would be if that node was actually useful in some way, that you didn’t need converters (and/or screencasts) to get simple lists of links from one application to another.


    Danny    Nov 28, 04:56 PM    #

commenting closed for this article

Google are too evil, or at the very least they are hypocrites Unorthodox rss use!!