|
November 30, 1999
Laboratory, Of The Future! Lab.01
is an interactive expo space developed by the good folks at Fork
Unstable Media.
Is there ever enough?
A profile of Too Much Coffee Man
creator Shannon
Wheeler is featured at *spark-online,
a smart new alterna-pop culture ezine.
November 28, 1999
Condemned to repeat it. Talking about the internet startup life
with my dad, he drew the analogy to the television industry in the
early 50s (A reportedly good chronicle of early TeeVee is The
Box.) I responded with, "And I hope we don't fuck things
up the way they did." He replied (and I paraphrase), "Of
course you'll fuck it up. You can't help it. You'll go after the
money, like they did, and that will be that. That's how things work
in capitalism." Sigh.
The tyranny of hierarchy.
Grappling with
the fundamental limitedness of hierarchy. Epinions.com, like many
sites, rests on a hierarchic classification of, well, everything.
And as the site grows, this taxonomy expands and its utility regresses
for two main reasons. First, such structures simply become unwieldy
to traverse. Second, different people classify things differently,
and as a hierarchy grows, more and more items will be placed in
ways dissonant to the user.
I guess my problem is
that hierarchy is a top-down approach that states, "This is
how things are. Obey."
What's a more bottom-up
approach? To let the data speak for itself. To let the users create
their own classification needs. To exploit relationships between
information. People ought to be able to slice through data as they
see fit. An example on Epinions is the gift area--it organizes some
products according to the type of person who would appreciate it,
crossing the site's item-type centered categories. But this is a
small step, and this cross-categorization is one we derived, so
it still has a bit of that top-down feel.
This relates to some
way-new interface concepts. Inxight's
Hyperbolic
Tree visualizes hierarchies. While it's kinda neat that you
can see a massive hierarchic space in a single window, it still
suffers from the burden of classification. And I think this limitation
will continue to prevent this interface from becoming popular.
On the other hand, PlumbDesign's
Thinkmap allows viewers to
decide what's useful. The Smithsonian
Without Walls interface has weightings for "era,"
and "theme," bringing forward the kinds of information
that you desire. The data is connected through relationships; among
other connections, chemistry set connects to toys connects to war
toys for boys connects to gender roles.
Hrf. That's all for now.
Still mulling. Searching around, I stumbled on this CHI
95 paper about collaborative filtering that bears on this discussion.
November 26, 1999
Snuffle. Sally
Tenpenny won't return my email.
We are amused.
Though I can't say I "get it," against my better judgment
I'm enjoying Daze of Our
Lives, state of the art 19th Century humor.
J'arrive! Philip
Greenspun points
developers who are "tempted to go the closed source commercial
software route" to my bvsucks page.
Philip's "Using
the Internet to Pick Up Babes and/or Hunks" is funny and
touching and, ultimately, envy-inducing.
Ever feel like you
were being watched? I was forwarded an email written by a BroadVision
employee, addressing point by point my bvsucks page. It begins "This
is a site that we have been aware of for some time now." With
this and Philip's note, I've updated the page.
(I won't go into details here, and bore those wise folks who aren't
web application developers. Though, those interested in corporate
cluelessness might find it entertaining.)
November 25, 1999
In the spirit of today's holiday comes this delectable online
comix blast from the past.
November 22, 1999
Faulkner
look out!
November 20, 1999
Swoon. I wonder if Sally
Tenpenny will go on a date with me.
You say "intuitive,"
and I say "familiar." Responding
to the Don Norman quote below, my dad illumined a rather significant
problem in interface design discourse, the misuse of the word "intuitive."
An often strived-for goal is an "intuitive interface,"
by which is meant that the interaction model for a device is immediately
apparent without training. However, only sentient beings can
be intuitive--it's
a quality of personal agency, not an attribute of an object.
Five years ago, Jef
Raskin (key architect of the original Macintosh) took this terminology
to task in the essay "Intuitive
Equals Familiar." (Go read it, it's good stuff.) Unfortunately,
designers haven't learned, and even the inestimable Don makes the
mistake of confusing intuitive (which, as Jef notes, ought
to be "intuitable") with familiar -- "things
that are intuitive almost always took thousands of hours of practice,"
and furthermore confuses intuition with a priori reasoning
(see the quote below from November 16). This is intelligible to
other designers mired in HCI discourse, but rightly struck my dad
as inherently contradictory.
Considering the semantic
predicament surrounding intuitive/intuition, and the foolhardiness
of striving for an immediately apparent interface anyway, it's probably
best to strike those notions from our design vocabulary.
Seek and ye shall
find. The britannica.com
search interface provides exemplary feedback. Query on "string
theory" and it returns a single prominent result on "superstring
theory." The same search on Encarta
returns a list of useless results. Britannica's database doubtlessly
uses intelligent keywording and related terms to provide the user
with the one result she most desires, which is dead-on most of the
time. Encarta most likely simply offers up a full-text index of
their entire database, leading to a large number of irrelevant results.
It's the old garbage in, garbage out maxim.
November 16, 1999
Time keeps on slipping... Eric
points me to the Whatever
Wall Clock, similar to the chronometry concept I wrote about
November 5.
I <heart> Don.
He opines,
Forget intuition.
It's highly overrated. And usually irrelevant, wrong, or both.
Blogrolling. Creative
Good's new goodexperience.com
blog focuses on ease-of-use news and information. And, as it's the
kind of person that I am, I already take issue with one thing Mark
has to say for interacting with palm devices: "The keyboard
is better for bit entry than Graffiti (the Pilot-specific printing
method), but it's not as good as voice recognition." I don't
take personal issue, but this assumption that voice interfaces are
The Wave of The Future puzzles me. Jeff Hawkins, creator of the
Palm Pilot and the new Handspring, was quoted
saying,
"Hawkins told
how his make-pretend method led him to conclude that voice recognition
will never be a good way to control computers -- a notion that
goes against current wisdom, including Microsoft's, which is sinking
millions of dollars into researching the endeavor.
Hawkins said
when he's sat around pretending to control his computer by voice
the experience is unsatisfactory and uncompelling.
It's not a
technical problem, he said, it's a problem of control, of having
the machine figure out what you're trying to tell it."
Jakob
agrees.
November 15, 1999
You say "Mer-cay-tuh," and I say "Mer-cah-tuh."
Again Lou and I tag-team
critique, this time assessing Mercata.com. Some of my column's
oomph got lost in editing, but I only have myself to blame (having
turned it in way after deadline.)
November 14, 1999
Ravenous. This whole internet startup thing is definitely skewing
my perception. You know how in the cartoons, when a character who
is really hungry looks at a friend, and the friend looks
like a big slab of steak with shoes, and the character starts salivating
at the prospect of devouring his friend? Well, I'm not going to
eat anyone, but dang near everyone I look at turns into a Potential
Hire in my eyes, causing me to worry that I'm being mercenary with
my friendships.
November 13, 1999
Point-and-click-illism.
I don't know how I've missed it, but Starry
Night is the neat-o keyword interface to Rhizome's
CONTENTBASE, programmed by Martin Wattenberg, who is also the proprietor
of bewitched.com (featuring
a different starry
night), and the creator of SmartMoney's award-winning MarketMap.
November 10, 1999
Get edumacated. Yesterday I should have mentioned that Richard
Anderson teaches a remarkable user-centered
design course which alighted me on the path I am today.
November 9, 1999
Lingual maneuvers. Gino pointed me to Olipop,
a delectable interactive gizmo. (Requires Flash.)
More on empathy.
The November 5 petermeme
on empathy spurred some thoughtful reactions,
much of it from folks on the CHI-Web mailing list. Richard
Anderson, usability guru nonpareil, pointed me to an
article, "Spark
Innovation Through Empathic Design." As I don't like to
pay for my nonpornographic web content (not that I like to pay for...
oh, never mind), I don't know what the article offers, but I did
find this
distillation of the main themes.
The tenets of empathic
design seem completely aligned with contextual
design, which, for my money, is the leading process for thinking
through design problems.
Karen McGrane, Information
Architecture Czarina for Razorfish, and, I believe, the person who
wrote "Reader
as User: Applying Interface Design Techniques to the Web,"
emailed me this thought:
There's some
sort of ... quality that seems to infuse people who make good
info designers. I don't know if I'd call it empathy or kindness
or what, but you're right to pick up on it. I can certainly pick
up on it within 5 minutes of meeting a candidate.
Many people
at Razorfish have commented that the info design group is consistently
made up of really good people -- thoughtful, kind, not competitive,
always willing to help out a new employee or someone who's swamped.
I'd always thought I was just lucky, or that I was just hiring
people I liked. But your email made me think that this quality
that makes someone a good team member might be inherent to what
makes someone a good user-centered designer.
This lead to a further
thoughtleap that non-empathic geeks become engineers, and empathic
geeks become information designers.
November 7, 1999
GUI Goodness. Steadfastly resisting the Blog Revolution, an
email from Leslie (Ms. Harpold
if you're nasty) points us to the Graphic
User Interface Timeline.
November 5, 1999
I miss my blog.
Nifty new Epinions.com
feature. For
all you bookworms and other media hounds--now you can review almost
any book, album,
or movie. The interface
needs a bit of work, but we're pleased as punch with the new functionality.
Forthwith, a fairly scattered
series of thought firings.
A Magazine We'd Like
To See: Martha Stewart Dying.
Tic toc so what? I'm
finding the concept of measured time losing meaning in my life.
If I were to design a clock for M & Co., there would be no numbers.
At the end of the hands, slowly circumnavigating the clockface,
would be the word "Whenever."
I feel remiss, as I haven't
been able to follow the series of articles on Salon including How
The Internet Ruined San Francisco, and I'm
The Enemy!. The latter piece personally upsets me, as it's the
article I wanted to write. In June I wrote a petermeme
titled Part of the
Problem, and I'd hoped to flesh that out into something
substantial. My mix of laziness and sudden workaholism is a detriment
to my writing output.
I
feel your pain. A while back, a fellow information architect
and I were discussing whether a mutual acquaintance could do the
type work we do. We agreed that while he was a very talented designer
and writer, a practiced aesthete, he lacked an essential quality
that any user advocate must have--empathy. A successful interaction
designer has to not simply suppress his own personality, but must
eagerly endeavor to understand the needs, desires, and methods of
his potential users. For better or worse, empathy is not a trainable
skill--you either have it or you don't.
And while this need for
empathy seems obvious, I've never seen it discussed in anything
I've read on user-centered design. As someone in the process of
hiring interaction designers, I'd love to have an empathy test I
could give potential candidates, not that I'd know where to start
in developing such a thing.
Continuing this thoughtwander,
what frustrates me so much about typical usability engineering is
a lack of empathy. Users aren't seen as people, but as subjects,
as an Other. Usability engineers don't endeavor to truly empathize
with the user's needs and desires. Instead they use accepted methodology
(time-to-completion studies, think-aloud user tests, heuristic evaluation)
to create an abstract model of the user that neglects the audience's
true humanity.
November 1, 1999
[too lazy to archive]
Hahahahaha! We've snagged another one! Rogers Cadenhead, best
known for Cruel Site of the Day
and the Drudge Retort, has
begun RefererLog, his new
blog. Where do you want to click today?
|