begin  quoting DJA as of Sat, May 28, 2005 at 10:06:16PM -0700:
[snip]
> My research: I looked at the green button and I don't care for it.
> M$ research: "Nah, I think the green looked better; go with the green".
> 
> Just because M$ did the research doesn't necessarily mean they came to 
> the best conclusions.

Even honest-to-goodness upfront well-meaning research can result in
a less-than-optimal UI. If 40% of the target population does best with
approach A, then you might have 60% finding approach A needlessly 
limiting and/or deceiving, but the best approach for that 60% might
be 29% for approach B, 29% for approach C, and 2% for approach D.

I happen to think that's how the world works, more or less. Thus,
"let's all copy just one interface" is an incredibly unhelpful thing
to do for most of the people, most of the time.

[MSWindows 1.0]
> At the same time that many of us still thought Quarterdeck's DesqView 
> was the kick-ass user interface on x86.

I knew a few fans of DeskView. It wasn't good enough to get me to give
up my Amiga, but hey, not much was.
 
[snip]
> People only talk up "Bang for the buck" when they can't afford what they 
> /really/ want. "Bang for the buck" applies when you need to compromise. 
> Compromising is not always the best decision.

My WRX was wonderful bang-for-the-buck. I would have _liked_ an STi, 
Lotus, or Maserati, but I couldn't afford those.  It would not be honest
of me to bad-mouth the choices I didn't go with -- especially in the
mistaken idea that it somehow justifies my decision.

Now, paying a lot more for a lot less (like all those people who bought
IBM PCs after the Amiga and Atari ST were on the scene) is often put
in the bang-for-buck category.  So I'm not _too_ quick to dismiss the
bang-for-your-buck argument, because it has its place.  But it's not
the only valid variable.

-Stewart "Plus, it's a crappy way to avoid buyer's remorse." Stremler

Attachment: pgp5OMAzuvUmr.pgp
Description: PGP signature

-- 
[email protected]
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list

Reply via email to