July 19, 2005
tag-only blogware etc. etc.
Sig's got a new plaything for all you geeks out there:
Tags, not Trees
Take my new tree-structure-free gizmo out for a test drive:
What is it?
Whatever you want it to be: Navigation-free website. Tags-only blog. Complicated-to-simple database. Active resource picker. Knowledge and learning base. File system. Whatever. An experiment for the heck of it.
Basically a different approach to organise data, finding data, and transferring knowledge.
An example of no-tree-structure-at-all. Anataxonomy in practice...
NB: This gizmo has no commercial application, or at least, it was built without a business model in mind. Like Sig says, he just built it for the heck of it.
What sayest thou?
[UPDATE:] Dennis in Sig's comment section referred to it as "Blog + Wiki= Collaborative Software v2.0"
[UPDATE:] Really good commentary from Doc Searls: "Politically, tags are The People's Directory."
[Disclosure: I work with Sig and his software company, Thingamy.]
Posted by hugh macleod at July 19, 2005 11:42 AM
Common guys, is that the best you can do? We've been doing that stuff as well as sucking in relevent tags from del.icio.us and flickr for over a year. Did i miss something?
Jim, sure, in del.icio.us you may use the "find" and add tag1 + tag2 and so forth, even directly to the URL.
"Tagwork" is a different interface - better or not is not up to me - we're trying too make the flipping around easier as a starter.
Then - it's done directly in the "web site, blog, data repository" and thus seeks to replace the navigation as per tree structure. del.icio.us still have navigation bars ;)
Actually, the objects in the experiment are not organised by anything else but "imprecise" tags - thus requiring a rather fast and easy-to-use interface to make it work. That's the experiment all about - see if it's practically feasible to replace tree structure organising... so I suspect more will follow this one.
Don't give up on us yet even if we may seem to brush against "what's been done before" every now and then :)
I would never give up on you guys - too much to learn from ya!
Cool! "Then - it's done directly in the "web site, blog, data repository" and thus seeks to replace the navigation as per tree structure. del.icio.us still have navigation bars ;)"
Our goals are the same: replace tree structures with a beautiful mess:-)
We use del.icio.us as an external bookmarking sys. We use the REST api in Ideascape to suck in tags/content from everywhere (diversity/independence), including the users, on a relevant tag + tag basis. We use AJAX a lot to let the user create their own structure, which makes for a richer user experience, one that compliments and enrichs the experience of finding and discovering ideas via navigation. Of course, tagclouds with a little AJAX work will for nav also.
What I am working on right now is refining a filtering technique to be able to identify an implicit community of interest and tap into its emergent group mind on del.icio.us and 3rd party services with accuracy that makes finding/discovering information on the long tail possible.
Don't ever stop with the experiments.
Funny, it seems to be in the air. I'm working on something very similar to what Sig just posted.
Oh yeah, and vaguely similar to what Jim has up. Nice.
If this is a conceptual preview of Thingamy, and Jim's AdvancingInsights is already out there, and my thing is in the planning/prototyping stage, there's bound to be a bunch of other people working on these kinds of things simultaneously.
There's obviously something about tags in the Zeitgeist.
Yeah, frosty, you're all 100% buzzword compliant.
Whoa, I've been slapped by Mr. Forward Markets! Pot, kettle.
Seriously though, one of the things that's attracted me to places like GapingVoid and Thingamy is the openness and civility of (most of) the comment threads. Except for the occasional Hugh-hater, the exchange of ideas here is usually very constructive.
Lest I seem too far off-topic here: it's got me thinking about whether tags should be augmented by some kind of ratings. Or maybe some kind of anti-tags.
Ideally (or not, but it's my direction) readers can tag things just like authors can. And ideally those tags get into the mix and are useful for everyone (delicious). But what happens when something has a tag (eg "buzzword") that for you is not only meaningless but potentially conflicting.
Say I actively use "buzzword" as a tag, but MarkN has tagged something "buzzword" that doesn't fit my sense of that. Maybe I should be able to exclude that tag from the object, and then it doesn't show up under "buzzword" for me anymore.
So far so simple. But how do I define the value to others of my having negated that tag for that object? Is that a useful factor, and if so how?
(thinking out loud...)
lmao you're so cool. thanks for being the perfect host and keeping the joint open all night.
Hey, I wanna join too!
(must been sleeping or whatever we do in offices...)
And Frosty, think we'll have something for you in version 3 (notice no humble 0.0.0.3 here) - not exactly what you suggest, but let me hint towards adding value to tags...
Say you've been tagged as "couch potato" by an ex-girlfriend - for us others to know that it was her doing that, and more about her would be cool. And add a lot of value to the tags :)
Kind of "source analysis" in History?
Hmm, source analysis is intriguing.
Something like a button to press for "tag details" - and then you get an area with that kind of thing (who's tagged something how, who's actively agreed/disagreed with the tag) - and that's probably where you'd want to make your adjustments as well.
Depending on how broad the user base is, it might even be nice to "ignore all tags by frosty's ex-girlfriend."
Still open to third party developers? My work is finally coming to a close, and I'll be kicking around for other applications.
What directions are you taking in promoting the concepts here?
One way to look at tagging is that it is simply the best that the likes of technorati and bloglines could come up with to facilitate content indexing and search within their space in a timely manner.
However, relative to the richness and robustness of Google's natural language processing, algorithmics and automation, manual tagging is an absurdly weak idea. The only rub with respect to Google is the lag between posting and indexing, an issue they are likely well on the way to addressing.
With respect to my inadvertent pissing-contest inducing labelling of the me too's as "buzzword compliant"...you'll get over it.
The valuable contibution Sig is making is the idea of tagging raw data, non-written content and other unstructured information as a way of facilitating its availability for consumption within an organization or between organizations.
now that "Kind of "source analysis" in History" sounds interesting - muti-faceted tags? But isn't the "source analysis" inherit in the object with time stamp and author? Tell me more...
MarkN, I think you're spot on about google catching up. The richness of tags is in the "tagging raw data, non-written content and other unstructured information as a way of facilitating its availability for consumption within an organization or between organizations."
you're on the track :)
The more I know about a "tag" the better - just like in history analysis - who wrote (tagged) it, when, what's the background and in particular, what's the relationship with the object/subject, any notes made when written (tagged)?
That would indeed increase the value of every tag and thus the "namespace" being the tagcloud for the object/subject...
Next version will have a "first try" at it of course :D