May 23, 2000
Snf. The Mojave
Desert Phone Booth has been removed.
May 22, 2000
Attack of the Goatees. And
in this photo,
I listen intently as Danny
O'Brien explains plans for world dominion.
May 21, 2000
Blur. The past
4-5 days have been a blur of sleeplessness, sociability, eating,
sightseeing, and head colds. What follows is, for the sake of my
memory, a chronicle of my trip to Copenhagen to attend the Reboot
conference. Let me start out by saying that I loved it, and owe
heaping scads of gratitude to our exceedingly gracious hosts. I'll
post a chronicle of my journey in the next couple of days.
UI patterns to consider. Often a bit too broad for my tastes.
Take a dip. Frank
points us to Serendip,
an interesting treasure trove of stuff related to complexity, biology,
science and nature, brain and behavior, etc. I'm too tired now to
root around it much, but it has potential.
May 16, 2000
Outta town. So
now I have an excuse for updating infrequently.
May 15, 2000
You must remember this.
Nick replies to yesterday's petermeme:
Yoz may have mentioned this to
you before, but some of my research deals with the Lakoffian concept
of mental spaces, and the use of spatial analogies to describe mental
processes. What I find interesting is that there's a specific point
of origin for this model (or at least I think that there is): it
emerges at the point where "modern society" begins, at the start
of the eighteenth century, and with the advent of a social and political
culture which regarded the self as individuated, and societies as
John Locke, I think, should be
thought of the original architect of information architecture, because
he wanted to think of the individual as if he were a Palladian building,
and to regard the memory as a kind of internal picture collection.
(Julian Dibbell's Feed
article on Jorn Barger has a point here.)
My point is that the Lockean view of memory is kind of like a camera
obscura: light forms shapes on a mental table (and I think this
is the proper way to think of a "tabula rasa", judging from contemporary
sources), and the tracings that the active memory makes can then
be stored and displayed, thematically, like a museum exhibit, or
stored in boxes like photographs and prints.
But yeah, if you're interested in origins of information architecture,
have a look at eighteenth-century picture collections, or at something
like the Pitt
Rivers museum in Oxford, one of those old cabinets of curiosity.
"Types", not regions.
What goes around...
urban design studies program that takes cues from information architecture.
Design, an online store featuring scrumptious modern industrial
and furniture design.
May 14, 2000
Thanks for the memories.
for Knocking at an Empty House, a collection of Bill Viola's
writings (from his notebooks and essays for magazines and exhibits--brilliant
stuff), there's a lot of discussion about memory, and some rather
approaches on how "concepts and terms for the manipulation
of solid objects are constantly used to describe thoughts, as in
'the back of my mind,' 'grasp an idea,' 'over
my head,' 'cling to beliefs,' 'a mental block,' and
so on."(Pg. 154) This spurred me to research the classical
notion of 'memory
palaces,' and other
What I've been wondering
is how memory spaces relation to information architecture. I'm not
interested in notions of a 3D cyberspace where you fly through information.
What tickles my brain is how information has no 'space' until it's
been consumed. Unlike a physical space, which exists before you
traverse it, the structure of an information space is built on-the-fly
in the mind's eye by the individual. For each and every person,
information exists only in memory.
May 12, 2000
What I saw at the Webbys. I feel obligated to write something
up about attending the Webby Awards. Piggy-backing on the nomination
of Metababy.com for best
personal site (!) allowed me entree to the event. Interviewed by
before the awards, I neglected to mention the name of my employer,
much to my colleagues chagrin. Merriam-Webster
Word Central provided the funniest acceptance speech, slowly
relaying 5 sesquipedalianisms. Google
co-founder Sergey Brin boogied to DJ Spooky's beats, wearing a hockey
uniform and roller blades. Bill
Viola (BILL VIOLA!) presented the SFMOMA Webby Award, and standing
near him at the after-party I was reduced to a giddy schoolgirl,
like, totally wanting to talk to him, but, like, afraid I'd sound
like such a DORK.
File under 'places
to check out when I have time.' Scott McCloud offers up his
list of 10
top comics web sites. [By way of Taylor,
whose blogs I really ought to read more often.]
Scott argues that stories
like Stan Lee's The 7th Portal
and Taylor's Star
are mislabelled as comics--while they feature comic conventions
such as panels and word balloons, they're animated endeavors, and
thus cartoons. Scott falls back on his precise definition of comics
as "Juxtaposed pictorial and other images in deliberate sequence,
intended to convey information and/or produce an aesthetic response
in the viewer."
Which makes me wonder
how Scott considers a feature like Fuckertown,
which mostly fits his definition of comics (though with some cinematic
flairs, like dissolves and pans, and the occasional moving bit),
but forces its own pace of reading on the user. Scott doesn't explicitly
mention user-control in Understanding Comics--though, I'd
consider that a key element of the comics experience is that the
user determines what they're reading, how they flip back and forth,
and the time they spend doing so. (This is particularly important
in visually dense titles like the original MAD comics, or The Watchmen
or Transmetropolitan, where you're always going back to remember
details, etc.) I'd argue that comics, in any form, must allow the
reader to control every aspect of moving through the story.
May 11, 2000
Smart readers! In support of my ongoing effort to not have to
write anything, Faisal has
sent in some interesting thoughts about dynamic information architecture:
Regarding your post about dynamic information architecture and
the analogies to spreadsheets and what they did for numerical modelling:
I've been thinking about this a lot, particularly in the area of
"concept exploration" (manipulating and working through variations
on a concept to find the best answer).
Most exploration interfaces are incredibly rigid: they give the
user a set of "axes" along which to move within information, allowing
the user to move along those and only those axes. This approach
is useful if the user wants the same axes as the interface uses,
but it is worse than useless if the user needs a different set of
One solution is to attempt to allow access to *all* possible axes.
This leads to different "types" of axes, depending on the kind of
information you're looking at. For instance, numerical values are
always on a scalar axis (cost goes from 0 to $x) whereas other kinds
of data fit into taxonomy heirarchies (species), relationship graphs
(network connections), etc.
In a past life I created an interface along these lines. It categorized
every piece of information we could find about an object and tagged
the information in a way that the end user could manipulate it and
move around within it:
e.g., for a car we would track size, color, speed, style of car
(within the style heirarchy), manufacturer (within the manufacturer
heirarchy), and several hundred other bits of information.
The key was to list every possible variable that would distinguish
a given object (concept/idea/etc) from any other object (obviously
all the concepts had to be at least similar enough within taxonomic
groupings that comparisons were meaningful). The tagging allowed
arbitrary queries ("show me all cars that are blue, are within the
'minivan' heirarchy, go 0-60 in less than 8 seconds, and are german"),
but the real power comes in the ability to "browse similar" by similarities
of the user's choosing (e.g. looking at a toyota land cruiser, decide
to move down the "size" scale, move up the "speed" scale, or look
at other cars made by toyota, etc.)
The upside to this approach is that users are often able to find
what they want by exploring in a direction that suits them. The
downside is that information overload is a problem: it's hard to
convey restraints on browsing and searching when there are more
than a small number of variables in play at once. It is also a challange
to identify all the "axes" (etc.) that people will want to move
around in and to present those axes in a usable fashion.
That all aside, I've been thinking more about how to apply this
all to discussions and group exploration. Most threaded mail/news
systems these days tend to fall down miserably when it comes to
breaking out concepts and threads. I'd like to see an idea discussion
board that had some way of adaptively tracking when threads had
split off, and showing them as such. The problem is that I have
no idea how to computationally notice that, and I suspect that the
"split" is going to be in the eye of the beholder (what you may
find to be a relevant new thread may be an integral issue to me).
May 10, 2000
Dig those electronic beats. Care of S.
Monkey, Esq. comes a pointer to the video
for 14 Zero Zero by Console.
Pleasing streaming for the masses.
May 09, 2000
Truly understanding the user. Choop
points us to Word Perhect,
and all I can say is go *right now* and play with this fabulous
word-processor art thingy that is very clever and ends up illuminating
some interesting issues in how we approach interfaces.
Denmark, anyone? So,
next week I'll be travelling to Copenhagen for the Reboot
conference. I've never been to Denmark. Any petermereaders out
there? Anyone else have suggestions for what to do? Email
Frivolous work. Michael
Schrage wrote great columns
for Packet (R.I.P. Packet,
one of the best original content publications in the Web's short
history), so I purchased Serious
Play, his book on the importance of lots of prototyping
in design processes. I suppose the gushing foreward by The Circle
Of Tom Peters ought to have warned me off, as the book proved to
be little more than a series of strung together quotes from Experts
blathering on and on. Schrage's thesis could have been sufficiently
presented in a 15-page-pamphlet, but I suppose you can't get a book
deal that way.
Still and all, there
are occasional insights to glean, and there's a whole chapter on
"The Spreadsheet Way of Life" that discusses how the world
of business got turned on its head by the invention of spreadsheet
technology--no longer spreadsheets merely kept track of what
was or is--folks could tinker with tons of different possibilities
for what could be. The impact of spreadsheet software simply cannot
be understated. Schrage sums up his thesis with
could statically track cash flows; software spreadsheets could dynamically
manage them. (Pg. 45)
And that, of course,
caused all kinds of bells to go off in my head. You good readers
know I've been ranting about the calculus of information, the need
to treat information as a changing, dynamic entity. Current information
architecture 'statically tracks' information. What we need are tools,
devices, architectures that allow for dynamic management--it leads
me to wonder if such a thing will cause as much of an upheaval as
spreadsheets did. Napster, its piss-poor interface notwithstanding,
suggests a direction for this notion. As, of course, does Launch.
I honestly don't know
if what I'm saying truly makes any sense, if it can actually be
implemented in any usable form. But it's fun to think about!
May 07, 2000
Hrm. That might have been the best episode of the X-Files all
Wheels in wheels.
I recently finished reading Paul Auster's New
York Trilogy, a brain-tingling collection of stories that
borrow conventions from detective tales, but take the genre to a
new metaphysical plane. Frankly, I was unprepared for the first
two works--I ripped through them as I would fiction, but upon coming
to the end, realized I ought to have savored them as good literature.
Auster's stories are puzzles with no solutions, but while reading
I found that simply tinkering with the ideas presented provided
plenty of enjoyment. Poking around, I found this
master's thesis on the trilogy, which does a good (if belabored
job), of discussing various motifs (Doppelgangers, 19th century
allusions, detective stories, etc.).
I don't care what the topic, I love it when subjects are discussed
systemically. Such as these theories of why
shopping carts are so big. If I were to return to school, I'd
study either cognitive science, or complexity
theory. At its essence, complexity theory states that all complex
systems are ruled by the same phenomena. And a recent finding from
complexity theory suggests that Dawkins' "selfish gene"
theory is wrong.
As mentioned here before,
The Tipping Point provides some interesting models for systemic
thinking, but does a poor job of tying them together. The best
critique I've found asks,
his entertaining and in part enlightening and educational foray
through all kinds of phenomena, one wonders what Gladwell has added
to the idea of the tipping point in its original realm, race. What
do Connectors, Mavens, and Salesmen have to do with transformations
in the racial composition of schools and neighborhoods? Very little."
May 03, 2000
I fucking hate the media. So I found out that television stations
aren't carrying the Epinions
Minivan TV ad (one of our best yet, though my favorite shot,
where she throws the keys, is hard to make out in RealVideo) because,
well, the reviewer in the ad details what a lemon her Plymouth
Grand Voyager is, and the stations are afraid of Chrysler's
wrath. I guess Epinions' money isn't good enough. And, of course,
they didn't even bother to air the ads and see if Chrysler reacted--they're
simply too chickenshit, and the economics of broadcast media engender
a remarkable chilling effect. Gah.
Annoy people! I've just signed up with imbot.com,
a service that will take your typewritten message and speak it to
someone, over the phone, in a charming robot voice. Feel free to
send me stuff: 415-516-6838. And, of course, since it's on the Web,
May 02, 2000
Doing it with Style. CNET provides excellent standards
for how to ensure the integrity of their style. This is probably
meant for internal use only, but it's nice that others can benefit
from their model. .