Welcome, Guest. Please Login
Tinderbox
  News:
IMPORTANT MESSAGE! This forum has now been replaced by a new forum at http://forum.eastgate.com and no further posting or member registration is allowed. The forum is still accessible via read-only access for reference purposes. If you wish to discuss content here, please use the new forum. N.B. - posting in the new forum requires a fresh registration in the new forum (sorry - member data can't be ported).
  HomeHelpSearchLogin  
 
Pages: 1
Send Topic Print
Reconciling visual & hierarchical thinking modes (Read 18734 times)
Rafter T. Sass
Full Member
*
Offline



Posts: 100
Burlington, VT
Reconciling visual & hierarchical thinking modes
Aug 01st, 2008, 9:27pm
 
The (perceived) inability to to move back and forth between a classical mindmap-style document and an outline that intuitively reflects the structure of the mindmap is, in my mind, a serious shortcoming of Tinderbox. If I could only have a complete, flexible, fully-functional graphic view of my whole document (which we can do very well in Tinderbox) AND have the visual hierarchy of links translate well into the nested hierarchies of the outline view (which we can't).

I use the hell out of Tbx, but for actually creating outputs from the information I'm manipulating, I'm left wanting for a way to translate visual structure into hierarchical structure.

I wonder - has anyone tried using some slick agents and the #linkedFrom and #linkedTo attributes (among others) to automate the generation of an alias-outline the shows the notes on a single-level mindmap as a nested hierarchy in outline view? What would this take?






Back to top
 
 
raughter   IP Logged
Mark Bernstein
YaBB Administrator
*
Offline

designer of
Tinderbox

Posts: 2871
Eastgate Systems, Inc.
Re: Reconciling visual & hierarchical thinking mod
Reply #1 - Aug 1st, 2008, 11:08pm
 
The implicit link hierarchy is only a hierarchy if you adhere to the very limited restrictions of Mindmapping(tm).  The problem is, if you use links in a less rigid and artificial way, you'll find the need for structures that don't map onto simply hierarchies at all.  

Even small projects can easily require multiple hierarchies.  

Yes, we've got some ideas about new visualizations.  And I'd love to know more about how you think this might work on realistic information sets.

It's a terrific research field, with several MA theses and at least one doctorate lying around waiting for someone to pick them up.  Seriously.
Back to top
 
« Last Edit: Aug 1st, 2008, 11:11pm by Mark Bernstein »  
WWW   IP Logged
Rafter T. Sass
Full Member
*
Offline



Posts: 100
Burlington, VT
Re: Reconciling visual & hierarchical thinking modes
Reply #2 - Aug 1st, 2008, 11:24pm
 
Point taken - it seems what I'm asking for is a way to operate Tbx in hobble-mode. The short-term gain (for folks like me who are just beginning to explore graphic knowledge representation) of having a module to translate a dumbed-down mapping mode back and forth into a hierarchical outline might be considerable - but also might be like welding expensive and clunky training wheels onto a light, nimble bike chassis.

Hmm. Maybe I can respond to the questions about realistic information sets after I start to put Tbx through it's paces, as I begin grad school in the Fall.

Any chance of news on new visualizations showing up in the dev peekhole?

By the by, just what IS the discipline that research into these questions falls into?



Back to top
 
 
raughter   IP Logged
Mark Bernstein
YaBB Administrator
*
Offline

designer of
Tinderbox

Posts: 2871
Eastgate Systems, Inc.
Re: Reconciling visual & hierarchical thinking modes
Reply #3 - Aug 2nd, 2008, 9:41am
 
A lot of the research in this area has always gone to the ACM Hypertext Conference, which has long been a welcoming bridge between the AI/Knowledge Representation community and the literary world -- both of which wrestle with representation.

Of course, visualization used to be the province of CHI.  My sense is that most of it has now wandered off to specialized workshops and new conferences like CKIM.

There's lots of other work as well in Personal Information Management, some in CSCW....
Back to top
 
 
WWW   IP Logged
Rafter T. Sass
Full Member
*
Offline



Posts: 100
Burlington, VT
Re: Reconciling visual & hierarchical thinking mod
Reply #4 - Aug 4th, 2008, 1:22pm
 
Mark, with the exception of AI, I don't recognize any of the acronyms.
Back to top
 
« Last Edit: Aug 4th, 2008, 1:23pm by Rafter T. Sass »  
raughter   IP Logged
Mark Bernstein
YaBB Administrator
*
Offline

designer of
Tinderbox

Posts: 2871
Eastgate Systems, Inc.
Re: Reconciling visual & hierarchical thinking modes
Reply #5 - Aug 4th, 2008, 4:07pm
 
ACM: Association of Computing Machinery. Along with IEEE, the major international organization for computer science research.

CHI: Computer-Human Interaction, a large annual ACM conference.

CKIM: a typo for CIKM, the ACM Conference on Information and Knowledge Management.

Hypertext: here, shorthand for the ACM Hypertext conference.  Many of the ideas in Tinderbox stem from research originally reported here.

CSCW: Computer-Supported Collaborative Work
Back to top
 
 
WWW   IP Logged
Rob Forsyth
Full Member
*
Offline



Posts: 63

Re: Reconciling visual & hierarchical thinking mod
Reply #6 - Aug 5th, 2008, 8:09am
 
Rafter has articulated well something I've felt for a while. Mark and I corresponded about what I think is a closely related issue a while back, namely leveraging the spatial organisation of the map view in outline view (i.e. the ordering of items in outline view would reflect proximity/position in map view). I gather Storyspace does this by default although I understand Mark's arguments for not having this behaviour hard-wired into TBX (the independence of the two views generally can be very powerful). On occasions I've found it useful however and have hand-kludged a solution based on agents sorting by xpos and ypos.

I vaguely remember playing with early versions of Apple's HotSauce (http://en.wikipedia.org/wiki/HotSauce) in the late 1990s which I think also touches on some of what Rafter's looking for, but how useful this would be in practice is I guess questionable
Back to top
 
« Last Edit: Aug 5th, 2008, 8:10am by Rob Forsyth »  
  IP Logged
Rafter T. Sass
Full Member
*
Offline



Posts: 100
Burlington, VT
Re: Reconciling visual & hierarchical thinking modes
Reply #7 - Aug 9th, 2008, 7:30pm
 
Now that I think about it, I'm really just looking for a slightly souped-up chart view - not as a replacement for the flexibility of map view, but as a separate tool for a certain stage of work.

I didn't think of it at first, because I never use it in it's current form.

Here are the desiderata that I think would give Chart View, and thus Tbx, a big boost in utility for non-savants (such as your humble narrator).

(1) Radial display (top of outline hierarchy at center of chart)
(2) Flexible size in icons/items - even, at minimum, auto-stretching to accommodate title text

Really, these (seemingly) minimal mods to Chart view would give Tbx 95% of the additional plug-n-play utility of concept-mapping apps like MindManager, without sacrificing any of the unique flexibility & power that are it's trademarks.

With this, I'd get to have my cake and eat it too. I'd get the dumbed-down hierarchical concept-mapping that I find so useful, have that translate live to Tbx outline, and still do brainstorming and sorting in a freeform map space with 9 flavors of linking.


Back to top
 
 
raughter   IP Logged
Rafter T. Sass
Full Member
*
Offline



Posts: 100
Burlington, VT
Re: Reconciling visual & hierarchical thinking modes
Reply #8 - Aug 14th, 2008, 12:47pm
 
Guess that makes this a feature request!

Back to top
 
 
raughter   IP Logged
Sumner Gerard
Full Member
*
Offline



Posts: 359

Re: Reconciling visual & hierarchical thinking modes
Reply #9 - Aug 15th, 2008, 6:51pm
 
Quote:
(2) Flexible size in icons/items - even, at minimum, auto-stretching to accommodate title text


The option for wider items in chart view definitely would help me too.  The chart view in its current form seems to truncate virtually everything, even though my titles aren't epistles and I have plenty of horizontal screen real estate.  As a result I can't use the chart view as much as I would like.

On the general subject of this thread, it seems I may be one of only a few who like chart view coordinated with outline view, but who do NOT like map view closely coordinated with the hierarchical views.

For me it is frustrating to have an effective map, showing relationships and links just as I like to see them, only to have the map "messed up" as hierarchical structure emerges in the other views.

I've tried manually creating aliases so I can keep the map and its visual links while changing the hierarchical structure in other views.  Though a bit of a kludge, that can be workable. But aliases, alas, always appear in italics.  And italics everywhere look strange on a map that I might want to show to others who don't know an alias from an elbow.

 


Back to top
 
 
  IP Logged
Rob Forsyth
Full Member
*
Offline



Posts: 63

Re: Reconciling visual & hierarchical thinking modes
Reply #10 - Oct 17th, 2008, 9:49am
 
This thread has gone quiet but I'm just seeing if we can pursue the general issue. Sumner is articulating my point in another way - the problem is the linkage between map view and the hierarchical views. You either need a way to view several "strata" of the map view simultaneously (better than the present double-click to open containers metaphor) - hence my previous mention of HotSauce, or (perhaps more simply) a decoupling that allows all notes irrespective of their position in the outline hierarchy to appear on a single "plane" in the map view?

Just my two penn'orth
Back to top
 
 
  IP Logged
AndyDent
Full Member
*
Offline

CTO of touchgr.am,

multi-platform

developer
/>

Posts: 7
Perth, Western Australia
Re: Reconciling visual & hierarchical thinking mod
Reply #11 - Jan 25th, 2009, 10:30pm
 
Using aliases, as discussed on the wiki, enables multiple map views that relate items from different levels in the hierarchy.

Unfortunately, the Tinderbox paradigm seems directly opposed to what you want, unlike traditional mindmap tools. Original items, rather than aliases, must all be siblings in order to appear on a map.

I think it would be possible, using agents, to generate a mindmap from an outline but not to go from the visual to generate the outline unless you started with a set of aliases in the first place. If your agent rearranged the outline of the items on the map, nested items would disappear.
Back to top
 
 
WWW   IP Logged
Rob Forsyth
Full Member
*
Offline



Posts: 63

Re: Reconciling visual & hierarchical thinking modes
Reply #12 - Mar 10th, 2009, 4:28am
 
Just thought I'd try to restart this thread, because it appears that with the capabilities available now in TBX 4.6 some of these ideas deserve revisiting. I'm thinking of what Mark B is calling "dashboarding" capabilities, i.e. that agents can now set the properties of the alias collected by the agent rather than the original. This could allow for example several different agents collecting the same notes but assigning  map-view Xpos and Ypos on the basis of different note attributes and creating multiple map views of the same original notes.

To me this seems a potentially powerful link from outline view to map view. I'm still not sure how well it will work the other way round (i.e. having an outline structure reflect shifting-around of notes you did in map view on the basis of perceived relationships)
Back to top
 
 
  IP Logged
Marcelo Mirage
Full Member
*
Offline



Posts: 133
Brazil
Re: Reconciling visual & hierarchical thinking modes
Reply #13 - Jul 29th, 2009, 12:33am
 
At first, I thought the semi-independence between outline and map view was a problem. I tried to work on both so both views would make sense. But, I'd like to point out that mindmaps and outlines are different in some respects, and now I appreciate what each has to offer.

Basically, there's one fundamental difference: Outlines suggest a logical sequence, from top to bottom. Items in mindmaps might or might not follow a sequence. Now, to read notes in sequence is not always meaningful. Siblings hierarchy, in other words, is not always desirable.

I've given up trying to reconcile map and outline view. Too much trouble for little gain. I tend to create notes that are meant to be read in outline, others to be read in map view. The notes for map view are usually in a container and I append a "MAP VIEW" in the container's title, so I can immediately recognize (when in Outline view) that the notes inside it are not meant to be read sequentially.

Of course, one could carefully synchronize both views, taking advantage of things like: A separator is invisible in Map View (and so are their children), as an adornment is invisible in Outline View. So, whatever notes you need to hide from one or the other can be placed inside a separator or be an adornment. That way you have control over exactly which notes will appear in outline view, which in map view.
Back to top
 
 
Marcelo Mirage Marcelo Mirage MMUmeda misantropov@yahoo.com   IP Logged
Pages: 1
Send Topic Print