Laptops today vary greatly in size, weight and purpose, but they all have one common origin: Alan Kay's Dynabook.
Kay, a former Xerox PARC computer scientist, drew up the idea of a portable computer in 1968, when computers still weighed over 100 pounds and ate punch cards. His definition of the perfect, portable computer was a very thin, highly dynamic device that weighed no more than two pounds.
Kay's original vision, which he called the "Dynabook," was never built. But it largely inspired the devices we now call laptops, although it's taken four decades to slim the tech down to the point where usable computers actually weigh as little as two pounds.
To honor his achievements, Mountain View's Computer History Museum on Wednesday will celebrate the 40th anniversary of the legendary
Dynabook.
Joining Kay in a panel discussion will be Mary Lou
Jepsen, the designer of One Laptop Per Child's XO notebook, as well as
Chuck Thacker, a former PARC researcher who co-invented Ethernet LAN.
In an interview with Wired.com, Kay shared the story behind how he hatched the Dynabook concept, and what challenges personal computers face as they continue to evolve.
Wired.com: So how did this all start?
__Alan Kay: __When I first had a full-fledged version of the Dynabook idea in 1968, I started to think about what portability really meant. I made cardboard models that I could fill with lead pellets to simulate sizes and weights. I defined
"portability" as "being able to carry something else too" and
"hand portability" as "being able to grab something else, too."
I
calculated that you needed about 1 million pixels to be really universal, and all of these combined to make a form factor that was very thin in one dimension, large enough for a real page display, and weighed about 2 pounds. There was some prejudging here because I had already made a drawing of two children learning science with these machines, so the final cardboard model wound up looking like the drawing, but with things sized and weighed out.
Later, at Xerox, one of the papers I wrote made the specs a little larger and heavier, when I was trying to convince Xerox execs to really take a shot at such a machine. But I think the original analysis still holds up pretty well.
However, my thoughts about an intimate personal computer were mostly of a service nature — that is, how could and should it act as an amplifier for human, especially child, endeavors? This is what led to quite a bit of UI, language and media design, some of which made it out to the commercial worlds in the 1980s. As a follow-up to the PARC
designs, the best thing I've seen in the right direction was Bill
Atkinson's Hypercard, which was not as general as our ideas but brilliantly filled in a number of very important properties that we were still vague about.
__Wired.com: __What do you think of netbooks? They're lightweight and small — pretty close to two pounds. Do they still need work before they can meet your definition of a
Dynabook?
__
Kay: __I'd like to think that they are finding a form factor and weight that fits human beings better, but I'm presuming that it is because many people use only a small part of what they could do on their larger machines, and much of what they do use computers for can be done through a browser or a few simple apps. So this would be somewhat similar to the limited uses of computing that fit into other even smaller devices such as phones and PDAs. If so, then this is more disappointing than something to be cheered about.
I cringe every time I use a browser for many reasons. The browser people had a chance to make a more integrated UI and functionality, but really did pretty much the opposite in almost all respects. But, because of the attraction, and even some real value of stuff on the internet, there is more pressure to do better. I would expect to see some real alternatives to the typical "bad defacto standard"
browsers we've had to put up with.
For example, suppose you actually have "real objects" as they were originally thought about. [These data objects] will have "ideas" about how they can display themselves and be edited, and [their appearances] can be moderated by various kinds of styles. The PARC UI took this approach and displayed the views from different objects in a 2.5-dimensional media layout space. Boundaries could be put around these views or not [to contain them within "windows" or not]. There were no applications, instead there were a highly integrated form of what today are called mashups. This was in the original PARC GUI!
There is much to be done here, and to even get back to a number of important integration and workflow ideas that were part of the PARC UI.
__Wired.com: __Are there any particular manufacturers that you think are heading in the right direction in terms of mobile devices?
__Kay: __All the ones I've seen have been spotty one way or another. The only one that has paid real attention to the screen is the OLPC XO, done by Mary Lou Jepson. It is otherwise a little too big, thick, etc. The service idea on it could be better, but it at least represents an attempt to rethink service, and has a few improvements on the standards.
The Amazon Kindle is kind of a subset of a Dynabook — too much of a subset. The screen is too small, it is not very capable of dynamics, the keyboard is poor, etc. But it does have several limited service ideas that are good. The next version of a
Kindle could be really exciting. The next versions of the e-Ink display are a much better size, they can be much more dynamic, the media range could be extended to what I called "Active Essays" years ago, etc.
__Wired.com: __What do you think will become of laptops in the not-too-distant future?
__Kay: __Does anything ever go away? In the early thinking about this whole space, one of the dominant factors is pixels and what to do with them. My thought back then was that we would move from a notebook device to head-mounted displays to get more effective pixels.
Another was the idea that it should be easier to make head-mounted displays than flatscreens (this is true, but virtually all of the engineering went into making flatscreens work). I still think that this will eventually happen.
What is really sad today is to see people sacrifice real functionality and real value for portability on tiny screens. The ominous side effect is that pretty much the only thing that can work on a tiny screen are the various things that work on television, and this is not a good thing for people.
__Wired.com: __What challenges do you think notebooks face as they continue to get more powerful and faster?
__Kay: __The biggest challenges are: A.) to really think up service ideas that actually help people; B.) how to get people to learn them, if they are actually new; and C.) how not to be pulled along by bad de facto standards.
The basic process is just the same as it was 40 years ago: define a viewing mechanism that presents above-threshold viewing angles, capacities, pixels, and define a service concept that can A.)
range over the enormous space of good things, and B.) that is integrated in concept so that "Simple things are simple, complex things are possible."
Photo and image courtesy of Viewpoints Research Institute
