[Sigia-l] Subdomain usability
ZO
listera at earthlink.net
Wed Jun 7 16:22:58 EDT 2006
Jonathan Baker-Bates:
> Any insight on why this happens?
I don't have any statistical evidence one way or the other. But I can
conjecture on the intent of users: convenience. For example:
Browser address box:
(Exact URL typed --> result page)
"maps/yahoo" --> http://www.maps.com/yahoo (Not a Yahoo property)
"maps/yahoo.com" --> http://www.maps.com/yahoo (Not a Yahoo property)
"yahoo/maps" --> http://www.yahoo.com/maps (Yahoo: "Sorry, the page you
requested was not found.")
"yahoo/maps.com" --> http://www.yahoo.com/maps (Yahoo: "Sorry, the page you
requested was not found.")
"yahoo.maps" --> server not found
"yahoo maps" --> server not found
"maps.yahoo" --> server not found
"maps yahoo" --> server not found
"yahoo.maps.com" --> server not found
"maps.yahoo.com" --> Yahoo Maps
In Google search box:
"maps yahoo" --> Yahoo! Maps and Driving Directions (#1 result)
"yahoo maps" --> Yahoo! Maps and Driving Directions (#1 result)
"yahoo/maps" --> Yahoo! Maps and Driving Directions (#1 result)
"maps/yahoo" --> Yahoo! Maps and Driving Directions (#1 result)
"maps.yahoo" --> Yahoo! Maps and Driving Directions (#1 result)
"yahoo.maps" --> Yahoo! Maps and Driving Directions (#1 result)
Which one would Average Joe rely on, DNS or Google?
For a company with a huge number of destinations like Yahoo, the DNS
permutation matrix requires a dedicated mapping application, with no
guaranteed way of total coverage. For the average user, the command line
interface of the browser address bar remains an enigma.
While we can marvel at how in fact DNS works as well it does, we can also be
befuddled as to why a better system hasn't emerged. It's no wonder that
people type *keywords* into Google's CLI as opposed to URLs into the
browser's CLI to find their destination. Google keyword mapping appears to
be far more "relevant," with only a few rules.
>From keywords to public-key schemas, there are a bunch alternatives to DNS,
with centrally-managed or distributed parsing processes. I'm hoping a saner
system will emerge to address a cascading architecture of related URIs with
more transparency, predictability and, of course, scalability, so we don't
have to do this dance forever.
----
Ziya
Usability > Simplify the Solution
Design > Simplify the Problem
More information about the Sigia-l
mailing list