</a>Possible additional features

Polyhedral Pattern Language: Software facilitation of emergence, representation and transformation of psycho-social organization (Part #5)

[Parts: First | Prev | Next | Last | All] [Links: To-K | From-K | From-Kx | Refs ]

There are clearly several options in this respect:

• additional features incorporated into subsequent releases of Stella
• separate modules (plug-ins) for optional use with Stella. In this case the approach is to get a module to work for a particular class of polyhedra, as with geodesics, and thereafter to build up plug-ins for various interesting classes of polyhedra.
• development of a distinct variant of Stella for the organization of data
• export to third party packages: Stella already provides for export in a variety of forms. The question is what might be the input constraints on other packages

It is important to recognize the prime function of Stella in considering such options. Such changes might simply complicate unnecessarily a well-honed application. One alternative is of course to allow users to select additional tool bars for facilities unrelated to the primary function -- such as some discussed below.

Labelling and images: As it currently functions, Stella ensures that colours, labels and images "stick" to surfaces to a very useful degree when rotated, etc. However, colours seem to stick over a wider variety of transformations -- including morphing -- than labels or images. Colours are also appropriately reflected in both 2D and 3D versions?

It would be convenient if labels and images applied to a face in 2D net representation could be preserved rather than lost when the image is folded into 3D. To what degree can labels and images be preserved through a greater variety of transformations

External links: These could be of various types:

• links to external web pages: Hotlinking from faces or vertices is not currently provided.
• links to sound files: An interesting possibility is to active a sound file associated with a particular face or vertex -- perhaps a key statement or commitment by a stakeholder
• links to other applications: These might include PPTs associated with progress reporting of an organization associated with a particular face or vertex.

All such links would currently require significant effort to export from Stella to VRML and might therefore be more appropriate in a separate application.

Identification of usable representations: With respect to on-the-fly input, one interesting feature would be to be able to call up (and/or select) a list of polyhedra with user-specified numeric properties, based on numbers of faces, edges, and/or vertices. Whilst polyhedra may be selected by type through a drop down menu, and many properties of any selected polyhedron are listed, there is no apparent means to search on such properties.

As an alternative, there is of course the Wikipedia List of uniform polyhedra, presented in tabular form, with images (in fact created using Stella) and with links to entries in Wikipedia on each. The table permits polyhedra to be identified by numbers or faces, edges or vertices. However the tabulation only includes a small subset of the polyhedra available through Stella. In Wikipedia the typology of uniform polyhedra is usefully described separately.

Thus if one had an organization with N major divisions, what polyhedra might one project (map) it on to -- in terms of faces, edges or vertices? Could the user run through a proposed list of possible "good fits" and make a selection?

How might one distinguish between primary features and secondary features, namely if each division had sections, how might these be attributed to smaller faces, etc? Or is the divisional level to be associated with groups of faces appropriately shaped and coloured?

Importing sets of linked entities: In the earlier online exploration, a set of linked entities was imported (effectively as output from a search process by keyword) and then fitted, according to a (crude) algorithm. The interface then allowed the user to intervene in the process (see image ***).

• The question is what kind of input format might be processed by such an algorithm -- clearly things could be made more or less sophisticated. That version put face or node labels into mouseover for VRML display and provided for hotlinking. Traces of this are still evident in the Java spring map version currently operational online.
• More interesting is the possibility to import an organizational nested hierarchy structure (line by line), or more generally a formatted network (links indicated to other nodes from each line), with a text label (for mouseover display) and possibly a URL (for hotlink). Input formats of this kind are common for network analysis and organization chart display programs
• Given the variety of entities that might be so input as a list (possibly of files), consideration could be given to distributing them across the available faces
  • randomly: as is currently done for colours in "rainbow mode" -- thereby extended to a list of images, sound files or labels, or possibly all three (CSV delimited) since they are not incompatible on a surface
  • ordered/clustered in some way (in the light of the following discussion on matching) -- raising the question of cognitively meaningful design metaphors for such distribution
  Already Stella provides for inputting a list of images to be randomly distributed to faces on a selected polyhedron.
• One of the possibilities, if the image has directionality and it is not feasible to turn it as the polyhedron is rotated (as is currently done with the label), is to provide for a means of turning the polyhedron so that the image on a face is upright
• Another possibility is to envisage a parallel split screen window (substituting for the net map, for example) in which items of information from such an imported list, and relating to each (numbered) face or vertex, could be listed in note form. The items so listed might include: name, geographical base, fund allocation, date, e-mail, etc. Whether this is open to editing in that mode, prior to subsequent export, could be considered. It might be easier to organize hotlinking such a parallel display.

With an easy input of a nested hierarchy of an organization structure, or array of nested strategies, this would enable a quick flip into a really integrative polyhedral overview of the initiative. Presumably outputs in a corresponding format might be made. One example of these approaches is the application Decision Explorer.

Matching and "fitting": Implicit in the previous point is the issue of how best to "fit" the structure to an appropriate polyhedron, or possibly a set of alternatives amongst which the user can choose or switch.

Clearly solutions vary greatly from simple to very difficult depending on how good a match is required. Issues include:

• Ability of the user to specify the goodness of fit, or accept a default
• Preference would clearly be for a polyhedron with as much symmetry as possible, but some numbers are easier to match to high-symmetry polyhedra than others
• Acceptance of blank faces where fitting cannot be achieved. Indeed one variant originally developed in VRML had either blank faces, or possibly an incomplete polyhedron where fitting could not continue -- or where the user specified a more complex polyhedron than was necessary
• Complexes of faces: With respect to nesting, clearly several approaches might be taken:
  • the smaller (secondary) face approach (mentioned above)
  • ability to allocate standard sets of faces (eg a pentagon surrounded by squares on edges and triangles at vertices) to portions of a hierarchy or to a network with particular local connectivity
  • representing each nesting by a different polyhedron, whereby clicking on a face would display a nested polyhedron of a lower hierarchical level

Clearly links should not be assigned to faces that share an edge, which would make a solution impossible most of the time.

Concentric polyhedra: The possibility of having concentric polyhedra (that could be displayed in wire-frame mode or with relatively high transparency) differently rotated. Seemingly this is possible, although trickier to arrange with Stella as it is.

A related possibility is to play with a light source from that centre such that, according to rotation of the concentric polyhedra, the light only "got through" certain "aligned" facets and not otherwise -- as an indication of consistency between different kinds of psycho-social structure (values, strategies, institutions and problems, for example).

[Parts: First | Prev | Next | Last | All] [Links: To-K | From-K | From-Kx | Refs ]