Reflecting on the Dynabook

I wanted to offer a couple of observations on the discussion topic for this week’s New Media Seminar, Personal Dynamic Media by Alan Kay and Adele Goldberg. The most striking thing about these early discussions of computer technology, in this case about personal computers and applications, is how prophetic they have proved to be. In particular I could readily identify modern analogues to the demonstration applications discussed in the article. From a computer architecture point of view Kay and Goldberg are focusing their Dynabook platform on desktop applications and personal computing, contrasting it with the shared time central computer systems more common in the ’70s. There are ¬†couple of points to take a look at here: where the interface lies, where the data resides, and where the processing happens. As the technologies behind interface, processing, communication, and storage have changed over time we’ve seen all these different areas of computing take place in different physical and logical locations. We’ve been on a path of doing more and more ‘on the desktop’ at the location of the computer interface but that has slowed and started to reverse with the rise of the Internet, ‘the cloud’, and generally increased communications speeds, ubiquity and reliability. Anecdotally, processing that I used to do on the desktop is now being shifted to ‘cloud’ based super computers and specialized back-end processing tasks are taking place on application specific servers. As a side note, I’m still not sure what ‘cloud’ means besides ‘resides somewhere other than the desktop’, it seems to be used in one sense as short hand for logically and physically distributed processing and storage; but more often than not it seems to mean ‘not on the desktop,’ which we used to just call ‘the network.’

The other point I wanted to mention has to do with the nature of application architecture, just as changing technologies have moved processing and storage alternating closer and further, the nature of applications have varied a great deal between paradigms of self-contained desktop programs, ¬†applet program components, web based applications, and device specific portable applications. I have a vague sense that the long dominance of self-contained desktop applications has more to do with a comfortable metaphor than any technical specification. The quick move toward and subsequent feint away from the WWW as application platform in favor of portable devices is the latest development and doesn’t completely make sense to me.

Finally, the question I always want to ask computer, network, and application architects is about optimum document storage. Are documents best stored with their parent applications or separately as ‘documents’ apart from their parent applications. The default hierarchical storage designs in most personal computers tend toward the latter, but that wasn’t always the case and with the move away from desktop storage and away from web applications I’m wondering if documents specifically associated with parent applications will become more prevalent. As a quick example, just about any new personal computer is going to come with a ‘Documents’ directory intended to store files, with applications and their support files hidden in sub directories in ‘other parts’ of the drive. Contrast that with applications that manage their own files like iTunes, or programs with specific file types that would not be intended to be used across applications.