Good thoughts by Ramana Rao. Of course technology will continue to improve and we will have the opportunities to concentrate on structuring information clearly.

I like the preface to that relates to some of your ponderings:

"Advances in science and commerce have often been characterized by inventions that allowed people to see things in new ways. Telescope, microscope, and oscilloscope are obvious instrument examples. But invented visual representations such as maps, statistical diagrams, and PERT Charts also qualify. Computers can combine both new visual representation, resulting in the emerging field of information visualization.

The foundational period of information visualization is now ending ... As this 15 year period draws to a close, there is a need for collecting together the results to date, organizing them, understanding the essence of this fieled, and providing materials for teaching."

I like the challenge we're offered. But typically, information visualzation texts have been written for other experts in the informaiton visualization field. I think we all have the opportunity to revisit this content from a user's perspective.

th

I noticed the thread, but did not bother to comment. I do not really have the time to deal with flame wars right now.
For me the lack of succes of Visualization for the past 10-15 years has more to do with culture and generation, than it has to do with lack of displays, killer apps or any other "device". As I see it Information Visualization is too hard to learn - both as a practice, and as a "reader". Most people have had enough trouble learning how to use a computer for text, and understanding the underlying metaphor of the web, to be able to deal with understanding multidimensionally visualised data. It will be a matter of passing generations before it will really catch on - also because generally speaking only the young generations really understand the need for "condensing" complex data - in my experience older generations simplify and dismiss information - rather than "digest" it - (but that is my own observation which I cannot substantiate with any emperical data). Information Visualization is like a new alphabeth in that sense - or like music videos perhaps. 20 years ago music videos were largely not understood by the parent generation. Now we do not even notice those visuals and editing practices that kept parents from even beginning to grasp what it was that their kids were watching. Information Visualization I believe is just too unfamiliar to most people who would not know a bar chart from a pie chart (and would gladly present a case of 100% chopped up in five "slices" - in a chart showing five pseudo 3d bars). just my 2 cent.
Gunnar

There does not seem to be much of a market for standalone "information visualization" tools. There are several reasons. One is that the web and Microsoft Office tools sort of dumbed down and ossified the user interface. The action currently is in data analysis (data mining and web search) and in basic communication (p2p, blogs, location-aware services, etc.) Another factor is that, so far, sophisticated data mining and vis tools are really appropriate for expert users, not a mass audience. Not too long ago a V.C. from a famous, wealthy company told me that he wanted to find infovis and related technologies that would make it possible for, say, non-technical workers to query and visualize large, semi-structured data sets assembled on-the-fly to answer complex business questions by point-and-click. My reaction was that we know how to crunch data and make pictures, but what's the point if the recipient doesn't understand what the numbers really mean? For example is it feasible to dumb-down the display of multidimensional data for a naive audience? I tend to doubt it. I think the idea of such a tool is great but it probably has to be aimed at an expert audience initially. I don't think in 2003/4 it is realistic to expert to find a large market for people who want to slice and dice multigigabyte datasets and solve the attendant data quality and interpretation problems.

We've had the best results here by working on vis tools (graphviz in particular) that we open sourced as to share with researchers and tool developers in other communities (e.g. software engineering, bio-informatics, network engineering, data mining) while assisting colleagues who built proprietary internal systems for problems like fraud management ("communities of interest"), software visualization of an executable specification language applied to VoIP services, display of various customer and backbone networks, and the like. We plan to continue with this model.