Tom Chi  

Why HCI is in the stone age

September 6th, 2005 by Tom Chi ::

An overview of some of the lingering vexations in interfaces today.

8 Responses to “Why HCI is in the stone age”
Greg wrote:

Interesting but totally off base.
The author assumes that most people are expert users and instantly learn the interfaces with keyboard shortcuts, subtle interactions and advanced functionality. In fact most people are not computer experts and often get confused by even simple applications. From reading this I would guess the writer has never actually watched a usability test with a beginner or elderly person trying to use a PC.

His comments about creating a socialist UI with no choices or preferences is plain stupid. People are not identical and they do different things with computers. People need choices to allow the interface to suite their specific needs.

Drew wrote:

1. I find that my screen is never, ever big enough to avoid hitting the corners by mistake. And the ratio of the number of times I activate a “handy” function unintentionally that gets in my way to the number of times I actually meant to activate Expose or my bloody screen saver or what have you is at least 10 to 1. Yes, the screen corners are extremely easy to hit. They’re too easy to hit. Whatever happens when I do it had better be harmless, or at least quick enough to cancel without seriously interrupting whatever I’d intended to do.

2. “Your search - a belly-barn shackle in the reunion of unjustified friends - did not match any documents.” Perhaps if I weren’t such a beginner I’d know what this means and where it came from.

3. This is a terrific point. I too wish resizing windows were easier and demanded less precision. I too feel sorry for the poor spacebar and its problem with void.

4. I can wait to monotonize all methods of interaction with the file system. First I’d just like to get around all the damn hierarchy and get the computer to realize there are only a handful of folders I care about on any given day.

5. I thought one-knob faucets were a revelation when I first encountered one. Imagine: now I didn’t have to guess how far I’d turned each knob, or how much cold vs. hot I’d need to make exactly the temperature I wanted. Taking a shower was as easy as (ta-da!) spatial memory: I thrust the knob in the direction I preferred with as much force as the amount of water I wanted. I have to assume there’s some usability lore about how these just SUCK but this particular user prefers them unconditionally. See response 1 in which usability wisdom is also refuted by (at least one person’s) empirical experience.

6. I’m all for spatial organization. Just remember that just as 1000 items in a folder is not “organized,” neither are 1000 folders on a desktop. We lay out the most important folders on our desks and the rest go into filing cabinets, sorted by some declarative method…because we aren’t using those folders every day and can’t develop the muscle memory to find them a year later. When we’re dealing with the volume of data we ask our computers to help us with, we need both methods to find things.

7. “User-centered design” is one of those phrases like “government of, by, and for the people”…only obvious in hindsight, and even then given a lot of lip service.

8. How about C. ::telepathic command::? I mean, if it’s a snap to do voice recognition and natural language processing AND it’s a great idea to perform a visual task using verbal commands, why not go all the way and demand psychic computers? It must be all the lazy engineers and dullard UI designers holding this innovation back with their grunts and flint knives.

And yeah: I still like one-knob faucets.

uurf wrote:

Did you actually read this post? I know it was slashdotted and all, but the comments on the post itself pretty well debunk his assertions. Rather than repeat them here, I’d suggest there’s a good reason why the author is anonymous, since these posts smack of someone new to their undergraduate coursework (”Fitts Law rules all!” “Radial Menus are clearly superior!”) mixed with the generally uninformed (seriously, “Alfred Einstein”?), and not worth revisiting.

Ron Zeno wrote:

Why HCI is in the stone age? Perhaps because people take such articles seriously. ;)

Kevin Cheng wrote:

FWIW, when Tom and I link stuff, it’s not necessarily an endorsement. It’s either, “hey this is an interesting interface” or, “let’s talk about this.” Shockingly, not everyone reads slashdot (I’m one of those neanderthals) so there are still links that people discover through us that they wouldn’t otherwise have ;)

And yeah, the comments to that blog do a pretty sound job.

Tom Chi wrote:

Whoops I guess I should have read this article… although the commentary brings up an interesting point. If the writer is just listing ‘textbook HCI’, and a lot of these textbook examples lead to suboptimal design… why are these things still in the textbook?

Robby Slaughter wrote:

Tom, the problem is really deeper than the examples here. People seem to still think “I know what’s best, so this is what I’ll inflict upon the world!” I never ceased to be amazed what you learn from user studies, and furthermore how much people resist the results and even conducting them in the first place.

HCI is hard in a unique way—not only is there a lot to learn, but the most valuable lesson is giving up your ego and yielding to the user.

Dave wrote:

Also, doesn’t MacOS X use corners? it’s the basis of Expose, right?

I found almost all the assertions meaningless and wrong. It sounded like someone who never actually watched someone use software in the field.

Leave a Reply

OK/Cancel is a comic strip collaboration co-written and co-illustrated by Kevin Cheng and Tom Chi. Our subject matter focuses on interfaces, good and bad and the people behind the industry of building interfaces - usability specialists, interaction designers, human-computer interaction (HCI) experts, industrial designers, etc. (Who Links Here) ?