User:Lieven Van Speybroeck/Reading/Theory/Bardini Bootstrapping

Notes on Thierry Bardini's Bootstrapping – chapt IV Inventing the Virtual User

1. On interface models
   • The phenomena of a person's mental model of the computer and the computer's understanding of the person are an important part of the interface next to its physical and sensory manifestations > problem of horrible recursion: both understandings would have to be included in the model, resulting in a never ending recursion
   • The interface as a join between computer and human > avoids the central issue of bringing it back to reality and representation
   • The interface as the representational space where the (interface)designer constructs the user > making the medium (~ computer) as transparant/invisible as possible by creating the illusion that the representation feels 'real'. Designing this illusion is designing the interface

   -> Make the users believe that what they do within the representation is what they within the real world

   • Virtual witnessing: the production in a reader's mind of such an image of an experimental scene that eliminates the necessity for either direct witness or replication

   -> creation of virtual users: products of the designer's own mental representations > designer and user at the same time

2. Engelbart's bootstrapping experiment
   • In search for the virtual user: the intelligence worker/knowledge worker

   -> Computer programmers as the ideal intelligence workers: autonomous & creative.

   -> Computer programmers as the template for constructing new kinds of users for new technology

   -> Computer programmers as a (his) crusade group

3. Black-Boxing the user
   • = The projection of a representation of the user via testing

   -> crucial in technology development: measuring the "virtual user" as envisioned by the designer with the real user. Seeking for inequalities or wrong estimations and making adjustments (ex.: the virtual user for the mouse kept changing during testing). During this process, the comparison with alternative devices that fullfil the same function is crucial (ex.: mouse versus light pen, tablet, joystick, etc)

   • Most important test results of searching 'the best pointing device' had to do more with the human factors involved in the design than with the performance of users with the different devices (select-operation hand, ease of the hand of gaining control of a selection device, fatigue effects of it's associated operating posture, ...)
   • Important characteristic of the virtual user for the design of the experiment was the degree of experience that real test objects had with various other devices.

   -> experienced users vs inexperienced users -> different testing procedures, different results

   -> conclusion was that only experienced user results were useful for an adequate assessment of the technology

   -> learning experience as a core point in the search of the virtual user: the virtual user envisioned for the mouse project already embodied a particular practice, namely the use of a REMOTE control instead of STRAIGHTFORWARDLY pointing device

   • Knee Control as an alternative to the mouse

   -> the drawbacks of the device for experienced users weighed heavier than its advantages for inexperienced users (for which the knee control had better results than the mouse)

   -> experienced users' results seemed to be more relevant for the ultimate development of the device

4. The gesture in interface design
   • For Engelbart, the mouse was not a project on its own, merely a part of making an interface. An interface for which he had in mind an input and feedback loop that was purely tactile, that depended completely (not on mouse or keyboard) on the gesture

   -> some kind of glove of finger caps that would record movement and typing mechanisms

   -> this required coevolution on the human side

   • Developing ways that computers could allow users to share and shape knowledge intersubjectively and collectively. -> NLS (ARC's oN-Line System) (~ applications/networks/...). Similar project: XANADU (first hypertext project) by Ted Nelson (was never finished/is still ongoing): Wikipedia on Xanadu
   • First presentation of NLS: "the mother of all demos" (1969)
   • Modular approach in the program design asked for different front-ends that serve multiple users. This approach needs a specific language and control structure -> a modal interface

   Problematic because <-> the way of working that users were used to: modelessly.

   By putting a bit of effort in it, people could easily learn the modular system. This results in an accommodation of people's existing working practices.

5. An accommodation of people's existing working practices wasn't the initial intention of Engelbart's search for interface design. It wasn't about making an 'easy to learn and use'-system, but more about improving quality and speed in the practice that experienced users had already achieved. The modal nature of NLS wasn't a problem, it was one of its core features.

Notes on Thierry Bardini's Bootstrapping – chapt VI The Arrival of the Real User

next up...