Copyright: 1992
Publisher: Apple Computer, Inc.
ISBN: 0-201-60842-1
Bruce "Tog" Tognazzini was Apple's "Human Interface Evangelist" at the time he wrote this book. His wry sense of humor and engaging writing style make this a fun read as well as informative. Tog was there in the formative years of the user interface of the Apple and Mac operating systems. His guidelines are foundational for those seeking to design user interfaces even today. The book's content and layout come from his monthly question and answer column in the Apple Direct magazine. While Tog doesn't seem to take himself too seriously (or maybe a bit too seriously?!) he definitely takes his work seriously.
Guidelines and Principles to live by
Throughout his book, Tog lays out a number of guidelines and principles as he is answering questions about interface design. The following is my personal reference of quotes, guidelines and other things to help me remember the things I found most useful in the book.
pg 27 - The keyboard interface should not dictate user design. This is because the keyboard interface is the "so-called power user" interface and tests have proven that mouse interaction is actually faster than keyboard interaction. Tog explains this by pointing out that using a mouse is a low-level cognitive action... you just point and click. Keyboard short cuts however require high-level cognitive action because you have to think about which keys to push and what they do.
pg 37 - Perception is stronger than reality. User perceptions can only be found through user testing.
pg 49 - A word is worth a thousand pictures.
pg 65 - Members of the development team are not the magical "average user".
pg 74 - The "average user" is a myth. The Air Force once conducted an experiement to find how many "average sized" cadets they had. They assembled 680 cadets in a court-yard and began calling off the average size of each article of clothing. Any cadet that did not wear the average size of say, the belt, had to leave the court yard. After only 5 articles of clothing, there were only 2 cadets left! After the 6th size was called out, there was only 1 left! Only one average cadet out of 680. Maybe there is no such thing as the "average user"!
pg 91 - Tog was heavily influenced by three different theories:
- Jung - Psychological types
- Shannon - Information theory
- Norman - Conceptual model
pg 130 - Don Norman on mental models:
- Mental models are incomplete
- People can't "run" mental models very well
- Mental models are unstable
- Mental models don't have firm boundaries (they often bleed over into other models)
- Mental models are unscientific (they often have superstitions in them)
- Mental models are parsimonious - people will often trade off extra physical manipulation for reduced mental complexity
pg 150 - Ockham's razor. You should strive to strip away all unecessary complexity
pg 152 - Follow some "real world" rules in your interfaces:
- Possbilities are limited in the real world so it is predictable
- World is populated with consistent, predictable objects
- Objects in the world can easily be perceived, discriminated and manipulated
pg 153 - If it ain't broke real bad... don't fix it.
pg 155 - Create new objects and appearances for new user behavior and for new resulting behaviors
pg 159 - Agents are "computer processes that act as guide, coach and as amanuensis" - Alan Kay (1990) [amanuensis]
pg 163 - Integrate typed input with pointing and other I/O channels. You can have the interface "talk back" to the user on the fly without always using dialogs!
pg 166 - Constrain agents to tasks that they are capable of. Make sure the user knows what a given agent can and cannot do... manage their expectations.
pg 189 - Porting code. Rethink your entire current application before porting, particularly the user interface. If you have a lot of "tools", you can probably replace them with a few well thought out tools that have several attributes and options.
pg 215 - Do not use the same control as a "status indicator" and an "input mechanism".
pg 217 - In modal dialog situations where the response can be guessed or inferred most of the time, used time closure dialogs. This avoids the "oops, I forgot to hit OK when I went away for lunch so my disk didn't get defraged over lunch" problem!
Cooper again?
I don't mean to pick on Alan Cooper... I really like his books... but I found yet another example of something he complains about in his book and yet there is a good reason for it. In About Face Cooper talks about how silly Microsoft is because if you drag a file from one folder to another, the behavior is inconsistent. If you drag it to a folder on the same computer, it moves the file. If you drag it to a folder on another device it copies it. Cooper points to this as a failure of Microsoft to be consistent in their user interface.
Tog mentions that this very thing came up in the design of the Mac interface. They implemented the same kind of logic (in reality Microsoft probably copied it from them!) The reason for it was simple... that's what the users expected to happen! As Tog says, "The most important consistency of all is consistency with the user expectations."
Bibliography
Here are a few things that I found reference to that I wanted to get back to:
[book] Graphic Design for Electronic Documents and User Interfaces (Acm Press Tutorial Series)
[book] Readings in Human-Computer Interaction: A Multidisciplinary Approach
[book] Lateral Thinking: Creativity Step by Step (Perennial Library)
[book] A Whack on the Side of the Head: How You Can Be More Creative
[paper] Usability Engineering at a Discount - Nielson
Conclusion
I certainly don't qualify for Mac Fanboy status... I sold the only Mac I owned last year. At the same time I recognize what a great job they have done creating a great user experience over the years. I admire the thinkers that have gone into (and come out of) the Apple machine. As such I think this is an incredible book and a must-read for anyone wanting to understand some of the groundwork that has been laid over the years for modern user interface design.