Welcome, Guest. Please Login or Register
Tinderbox
  News:
Tinderbox 6.5.0 is now available! (OS 10.10+) http://www.eastgate.com/Tinderbox
  HomeHelpSearchLoginRegister  
 
Pages: 1
Send Topic Print
Outline organization vs. Map organization (Read 10 times)
Scott Laird
Ex Member




Outline organization vs. Map organization
Dec 31st, 1969, 7:00pm
 
I'm new to TB, and I'm hitting a bit of a wall.  I'm not sure if it's just a mental model problem, or if this is a missing feature.

I'm trying to use TB to help design a small piece of software for work.  I don't fully understand all of the requirements yet, so I'm entering requirements as notes in one outline hierarchy, modeling bits of code with notes ("do this" in one note, then "do something else" in the next) in another place in the hierarchy, and making notes on datastructures in another place in the hierarchy.  I'm adding questions that aren't fully met by bits of design as more notes and linking them to the places that they're applicable to. I'm intending to use agents to link bits of design to places that refer to the design, but that's still a day or so away.

The problem that I'm hitting is that I can't seem to lay things out so they're useful in both map views and outline views.  For instance, I have a set of requirements down a level or two in the outline view and a use case in another place in the outline view, plus a couple open questions in another place in the outline.  I'd really like to be able to lay out a map where all of the questions, use cases, and requirements are visible at the same time and the links between them are visible, but this doesn't appear to be possible without building tons of aliases.  I'd really like to treat outline hierarchies more like adornments, where the sub-nodes are easily visible.  All I can seem to do is get a bunch of dinky little boxes inside of the upper-level note's map entry.  This makes it impossible to visualize the links that I'm creating.

How should I approach this?  Should I put everything on one outline level?  Should I use boatloads of aliases and/or an agent?  Is there an attribute that I can set to make the map work better for me?
Back to top
 
 
  IP Logged
Bernstein
YaBB Administrator
*
Offline



Posts: 399

Re: Outline organization vs. Map organization
Reply #1 - Dec 31st, 1969, 7:00pm
 
It's possible that you really want to make two documents, each with a radically different organization!

Using aliases liberally might help here.  So, too, might simply linking related things, facilitating navigation across the outline.
Back to top
 
 
WWW eastgatesystems@mac.com   IP Logged
sjhorn
Ex Member




Re: Outline organization vs. Map organization
Reply #2 - Dec 31st, 1969, 7:00pm
 
Welcome aboard, Scott. I must say that I tend to agree with you, and this question comes up every now and then.

Most of the time, the way I use TinderBox is basically hierarchical, i.e., as an outliner. But I often find that I would like to have a Map view that can either show that hierarchy, or link bits of different parts of the hierarchy using adornments. In other words, several notes from different parts of the outline hierarchy may need to be grouped together (under a different category), but they may not all be at the same level of the outline. If they are not, they can not be visible in the same Map. Tinderbox won't allow this this unless you use aliases or copies of notes (in other words you really are creating two documents in one).

With all due respect, Mark, I think this is a limitation, although I can see that programming a Map view that does what I would like would probably be very difficult. I am not sure what the solution is but I think Scott states the problem well: "The problem that I'm hitting is that I can't seem to lay things out so they're useful in both map views and outline views." The key word is useful.

Best wishes for the holidays to all.
Simon Horn
Back to top
 
 
  IP Logged
Dave Garbutt
Ex Member




Re: Outline organization vs. Map organization
Reply #3 - Dec 31st, 1969, 7:00pm
 
Hi,

I am trying to do something very similar for a project I am working on. There are over 200 user requirements and it is hard to manage.
Currently I have a hierarchy but I want more links -
like linking each URS to a set of possible solutions. I want to show these 'solution maps' to people and look at costs, advantages etc.

It seems to me that this characteristic of TB is a limitation. I don't know another or better solution but that doesn't help - it illustrates how TB has liberated my thinking and then tripped me up.

I recall a thread along these lines in September(?) about showing the prototype links in MAP view, and having an option to suppress them.
This problem is a generalisation of the other. And a possible solution is the same - have options (that are properties of the map view not the document) that determine if hierarchical links are shown, or not, in the map view.  
Another option might be to display hierarchical links as trees (from top, from base) or as centered on the top item (giving a look similar to a mind map).

Allowing user selection of the links-types to be shown would be a powerful tool. This means that all kinds of different diagramming views are possible, without making aliases. This seems to me  Wink an elegant solution.

It also perhaps means that the template functions ^child of^ and ^contains^ could have a different interpretation based on the kind of link that exists.

Further allowing this kind of duality (well, multiplicity) of views makes TB a truly link-based tool rather than an hierarchical 'outliner'

Dave
Back to top
 
 
  IP Logged
Scott Laird
Ex Member




Re: Outline organization vs. Map organization
Reply #4 - Dec 31st, 1969, 7:00pm
 
Another possible solution would be an attribute on notes that would make them act more like adornments--when it's set on a note, and the note has children, then the children will be visible in a map view whenever the parent is visible.  This *should* just change the drawing logic a bit without being a major change to the code.
Back to top
 
 
  IP Logged
seolds
Ex Member




Re: Outline organization vs. Map organization
Reply #5 - Dec 31st, 1969, 7:00pm
 
Scott,
Could you explain this a little more? My solution has been to make the whole map/outline a single level with adornments linking pieces together. But then waht. I'll post a pic of my map - any suggestions would be great.
-Shelley
Back to top
 
 
  IP Logged
seolds
Ex Member




Re: Outline organization vs. Map organization
Reply #6 - Dec 31st, 1969, 7:00pm
 
OK, I've uploaded an image of my map at http://see.gsfc.nasa.gov/earthgazer/JDL/structure/ESEDLtboxMap.jpg

So, I have all of the notes at the same level of the hierarchy. I'm a visual person, which is why I bought Tinderbox. But I plan to place the note contents online with like items within a hierarchy.

Ideally for the outline I would like like colored items to be inside of a single folder - e.g. metadata items inside of the metadata folder. But then in map view, I can't read the names of the notes within the main note.

Any suggestions? If you suggest Agents, please be nice and describe your idea completely, I'm new at this.

Thank you,
Shelley
Back to top
 
 
  IP Logged
Bernstein
YaBB Administrator
*
Offline



Posts: 399

Re: Outline organization vs. Map organization
Reply #7 - Dec 31st, 1969, 7:00pm
 
This is a very nice map!  Wonderful!

A common scenario in the research literature that is the foundation of Tinderbox is the idea of emergent structure.  At early stages, you don't know how to organize: a freeform map seems best.  Items often move often, sometimes radically, and experiments with secondary organization (color patterns, shapes, adornments) are common.

Later, parts of the project become better understood.  The mao changes less, and you refer less often to some of of those parts.  At that point, it makes sense to rearrange the static parts into a deeper hierarchy with more formal organization.

Later still, of course, unexpected things arise, and you'll use the map space and complexity you just reclaimed to represent those unexpected things.
Back to top
 
 
WWW eastgatesystems@mac.com   IP Logged
seolds
Ex Member




Re: Outline organization vs. Map organization
Reply #8 - Dec 31st, 1969, 7:00pm
 
Thanks Mark - to add a slightly different perspective: I know that I want to place the like colored stuff inside containers (which are just notes, right?) but being a visually based person and also needing to share this with many people who are unfamiliar with digital libraries or content management, I need to keep these little notes visible. What to do? I've thought of making two T-box documents, one for looks and one for building the website. But I think I can build the hieracrchy through Agents and attributes.

Now a practical question.  If you look at the right side of the map, I have vertical columns(notes) and horizontal rows (adornments) that interect, then there are little notes that belong in the intersection. Again, I imagine Agents or Links would work - are there any advantages/disadvantages to either method?

I'd really like to see some links to this research as part of my job has been to help people with their metadata, taxonomy, and website development. Having some research to back my insistance of not creating deep structure too early has met with resistance. What they find is that some items need to be present or linked into multiple stacks/stovepipes and then at some critical mass of links, the item should become its own top level.
-Shelley

Back to top
 
 
  IP Logged
Bernstein
YaBB Administrator
*
Offline



Posts: 399

Re: Outline organization vs. Map organization
Reply #9 - Dec 31st, 1969, 7:00pm
 
The relavant buzz-word in the research literature of the field (which is often called VISUAL HYPERTEXT) is premature commitment.  You'll find this widely discussed throughout work on knowledge representation, argumentation, and hypertext rhetoric.

The top people in this area are probably Cathy Marshall (whose papers have chiefly come from Xerox PARC and Fuji FX but who is now at Microsoft) and Frank Shipman (now at Texas A&M), and the classic early polemic is their joint paper, "Formality Considered Harmful".  There have been two recent workshops in the area at the annual ACM Hypertext Conference, and you'll find many key references in those Proceedings.  

I would especially recommend Marshall et al on Information Triage, which describes one approach they observed in ethnographic studies to the problem you describe: how to you have structured information without information hiding?  In essence, they build semi-structured clusters with a central note and a halo of tiny subnotes, arranged in a predictable layout.  Over time, it may be convenient to bring this halo inside the collection, but the intermediate state (visual subordination) provides some useful flexibility.

That said, the problems you describe rest at the edge of profound and difficult questions in knowledge representation and indeed epistemology.  The key point (and one your critics may overlook) is that early formalization works well in problem domains you understand fairly well, where classification errors are likely to be small (so things arent very far from where they belong) and rare (so the structure feels meaningful).  This is often the case for an expert, doing incremental research in their own field.  It's not a good approach, though, when you really have little idea where the research will lead you, or what to put where.  Some fields (archaeology, for instance) have notorious problems with premature commitment; if you look for a field where people are notorious for publishing late, where they often have mounds of raw data waiting to be understood, you'll likely find one where premature formalization causes lots of trouble,
Back to top
 
 
WWW eastgatesystems@mac.com   IP Logged
Pages: 1
Send Topic Print