Alan Cooper
Feedback and Commentary on the Interview

Interaction design is not a person, it is a task.

The statement "The no.1 thing that you can do to create a better product is to bring the Interaction Designer in early." is short sighted, and clouded by a "disciplinistic" attitude. The no. 1 thing that you can do to create a better product is to begin and end with an integrated, multi-disciplined approach.

Nathan Joseph Fortin,
Senior Interface Designer

David,

Excellent interview! However, it doesn't answer the one burning question I have about Alan Cooper: If users hate products with complicated, unnecessary features, why does Cooper buy all the high-end, feature-ladened crap that he does? :)

Good work,

Everett

David

Nice job with uidesign.net.

I found your link on the epssinfosite, and it's loaded with what I consider useful information. My only negative comment is that you need a copy editor - in your interview with Alan Cooper I saw too many incorrect usages of its and there (instead of it's and they're). At the risk of being flip, can I suggest you have a disinterested 3rd party proofread your articles prior to publication. Your info is great, but you undermine your message when it's got typos and usage errors.

Anyway, those are my two cents. Keep up the good work.

Cheers,

Josh Geller

Hi David,

This was a good interview. I enjoyed reading it.

Thanks David!

..brad

Hi David

"You might have a real world situation where you have someone who walks into a library and searches based on a DD System number and then wants to see a list of related books. The query system fundamentally disallows this.

There is this whole class of human behavior which is not supported by the relational database query paradigm."

I would argue strongly, that the problem (which is pretty much ubiquitous and very real) is in the classificatory system, and has nothing to do with the relational query paradigm. The DDC predates (1873) the computer, and is a method for physically locating something, in this case a book, or more precisely a physical copy of a book. Within a database, physical location is meaningless, and there are any number of ways to locate something. And bear in mind, the problem is locating a book in the abstract, not a physical copy of the book.

I can recall on a couple of occasions trying to persuade people to dispense with a classification system, because it served no real purpose in a computer system, but ran into a psychological/emotional attachment. Imagine explaining to a librarian that the DDC system (or ISDN/Library of Congress) no longer serves a purpose and they should get rid of it!

The internet search problem, alluded to earlier in the interview, is caused by searches based on keyword incidence.

Most people in most situations want to search based on semantic content (what is this book about) and non-semantic properties (suitability for different audiences, etc.). Someone must define these properties for each book, and once they have, a relational database can easily be constructed to support this. Whether this is efficient/scalable or not, on any particular relational DBMS implementation, is a whole separate discussion.

Phil Bradley