Tinderbox User-to-User Forum (for formal tech support please email: info@eastgate.com)
Tinderbox Users >> Tinderbox applications >> Learning Curve

Message started by Mark Bernstein on Mar 7th, 2009, 1:56pm

Title: Learning Curve
Post by Mark Bernstein on Mar 7th, 2009, 1:56pm

A nifty new page in the Tinderbox Wiki:

A very nice discussion of ideas from Doug Miller, Ken Tompkins, Doug Vornov, and others.  Highly recommended.

I should note here that this discussion applies more broadly to nearly all information and knowledge management technology, and not just Tinderbox. I’m surrounded by friends, family members, and colleagues who see no connection between their computers, mobile phones, the InterWeb?, databases they access, e-mail, and their own internal thought-space. To them, all these things are more or less separate and distinct entities, and connections between them are seen as difficult and occasional. They are mired in metaphors that aren’t wholly appropriate for the technologies they use. Computers are typewriters and calculators. Mobile phones are just that – phones. E-mail is like a paper letter or card. Networks are, well, just plain black magic.

Title: Re: Learning Curve
Post by JM on Mar 10th, 2009, 7:44pm

Food for thought  8-)

This is an extremely interesting text, thanks for that!

Title: Re: Learning Curve
Post by peter lindsay on Mar 22nd, 2009, 5:09am

well...networks *are* black magic, ain't they ?

Title: Re: Learning Curve
Post by Jan on May 10th, 2009, 1:30pm

Oh, yes.  The ref'd wiki page is spot on when it says:  

The more information contained within a Tinderbox document, the more useful that document becomes.... Early on, the new user has little data entered into the program ....  Functionality at that point isn’t perhaps as obvious as it will be later...

I'm right at this exact spot today, making my 3rd or 4th try in a dozen or so years at grokking the technology to see how it would support my need to tie bits of computer-stored info together.  I want to link text stuff and URLS from the web and my emails together with images I'm working on -- a sort of project status application.  I know how I would use Filemaker to do this but am trying to determine whether Tinderbox would integrate more easily with resources and applications outside itself. In particular, info and pixels extracted from images by scripts rather than primarily text.

I went through the tutorial step by step by step, looking each time for what had happened in the definitions.  Then I tried some baby step note and structure creation with some of my own data.  Went back and tried the tutorial again.  Several rounds, having to quit Tinderbox each time.  Maybe I could get it ....  *BUT* ... what I couldn't do was test it with any knowledge set I understood that was both large enough and exemplifying the structure stuff.

This was my major problem:
  • The download-as-demo is limited to 30 notes.  (Complex applications nowadays often offer 30 days trial instead.)  With such scant data, it is hard to get a sense of  how the app is going to feel with a larger data set size and what the tools are for managing it all.

My pedagogical point, however, is that learning something complicated is easier when you can modify an existing complex structure rather than trying to get all by yourself from learning about simple note creation to the real fundamentals, which are linking and agents and probably more that I haven't encountered yet.  Aside from the data, one has also to assess whatever tools support surveying and maintaining one's complex data structure.  (Say, does Tinderbox support forgetting?)

From my reading, it didn't look as if one could distribute a structure without its contents.  e.g. can't have a prototype without a note to base it on.  (Maybe I'm wrong in thinking of prototypes as being like templates.  Be kind, I've been looking at this for all of 4 or 5 hours!)

Suppose a learner could get a full source structure for a work problem *similar to their own problem* to work with -- perhaps an elaborate structure that was sparsely populated.  Then there would be some further scaffolding for climbing the learning curve.  (Feel free to tell me it exists already and I'll check it out.)  Right now, there seems to be a huge conceptual gulf between the help tutorial and understanding the material in the wiki.


Title: Re: Learning Curve
Post by Mark Bernstein on May 11th, 2009, 11:50am

A prototype *is* a note.  It's just like any other note.  It might have started life as a grocery list.

But another note can use that note as its exemplar.  Next week, instead of starting your grocery list from scratch, you might say, "OK: make me a grocery list just like last week's -- and then I'll tell you what's different!"  

It's nothing more complicated than that. (But it's extremely powerful.  The AI/knowledge engineering people have been using the technique for about a decade...)

You'll find a few good frameworks in the Tinderbox File Exchange.  And there are more on the Tinderbox Weekend Remote Membership CD. (I'd *strongly* recommend attending a weekend rather than just poring over the CD.  The last weekend was terrific, with examples ranging from Parliament to personnel reviews to project dashvboards to plotting a novel.  Next one: San Francisco this November)

Title: Re: Learning Curve
Post by Mark Anderson on May 11th, 2009, 1:35pm

Whilst it is less how-to than the wiki and such resources, you might want to have a browser window open on my aTbRef resource. It should help you find the right knobs and switches as you explore Tinderbox.

Tinderbox User-to-User Forum (for formal tech support please email: info@eastgate.com) » Powered by YaBB 2.2.1!
YaBB © 2000-2008. All Rights Reserved.