Hi Mimi,
I recall a UI designer, whose name I can't remember, once said that you
should evaluate a design by looking at 3 criteria: clarity, efficiency,
and
aesthetics, and in that order. So looking at this design, I have the
following thoughts:
Clarity: Clarity is supposedly enhanced if the design is familiar,
because
it reminds you of something in the real world or follows a familiar
convention for computer users. Although some users might recognize the
eye icon in the design, I don't think that knowlege will help them
understand
it's function, e.g. include the collection in the summary view, in the
same way
a check box might. It also has the disadvantage of not following a
familiar
computer convention.
Efficiency: Here the design does very well. It packs checkabliity,
color and
iconness in 16X16 pixels and doesn't impose any unnecessary extra steps
to
get your job done.
Aesthetics: The idea here is how does it look. This can be in the eye
of the
beholder, but in general, some designs just stand out from others.
It's why
Macintosh UI looks better than Windows UI. For example, most people
think
the old "burn icon" in iTunes, the OSX "Genie Animation" and
www.imaginationcubed.com
rank high. In this respect, I think the proposed
design is hampered by too few pixels to work with, and the animation
will
likely seem unnecessarily distracting to users.
Finally, there is the question of implementation, which impacts not the
quality of the design, but the cost of the project. The proposed design
would be more expensive to implement because it doesn't take advantage
of a
class library. (Albeit, since that last design was also completely
custom
UI, we now have a more flexible implementation that makes it easier to
try
out other designs). But, in the end, I think this may not be the most
important issue to spend our limited implementation dollars on.
Sometimes the simplest solution is the best solution, and I think this
might
be one of those times. If you want an icon, check box, and color, I
suspect
that the simple solution of an icon and colored toggle control, (e.g.
the
Thunder bird "read" toggle control) is probably best. To me, on
clarity,
efficiency and aesthetics it's as good or better than the proposed
design in
all ways except one: it takes more screen space. However, I think in
this
case, screen real estate is a fine resource to use. I remember at NeXT
when
Steve wanted to make large icons, there was resistance at first, but
after
seeing the result, both on NEXTSTEP and now on the Mac, I think it's
best to
not compact too much design in a small space.
John
Mimi Yin wrote:
After confirming with David that we can use animated gifs
in Widgets, I've put together an animated version of the rollover.
In addition, I've changed the checkmark to an eye (the idea being that
if we're not going to use standard widgets, it might be less confusing
to use a completely different metaphor).
I've also created a generic collection icon (it's supposed to look
like a tag, as in a price tag), but right now it mostly looks like an
ipod mini.
Again, please drag into a browser to see animation. Still needs visual
tweaking, but the concept is there.
Mimi
On Oct 27, 2005, at 3:42 PM, Davor Cubranic wrote:
Mimi Yin wrote:
I think this is something that mostly needs
user observations. I agree it's a sticky issue, but it was the
solution we decided to try. We went through a process. We considered
many of the suggestions that have been raised and we settled on the
"checkbox on rollover" solution as one worth testing in the field.
[...]
We understood it was unconventional, but we wanted to leave options
open to unconventional solutions
[...]
I think we should give it a chance to work or not work as well as
ourselves an opportunity to observe people using the sidebar and
collect feedback from a broad spectrum of users (especially ones that
don't think about how software is designed ;o) before deciding one
way or the other. More comments in line...
That's a worthwhile idea and I'm fine with that. What might help is to
adopt unconventional solutions in the rest of the UI. (Not that any
are coming to my mind right now though.) It's like UI in computer
games: game designers are free to adopt highly unconventional, but
also highly effective and often immediately usable, interfaces because
it's immediately obvious that they are not following standard desktop
application guidelines. But when everything else in the UI looked like
standard widgets, the icon's behaviour on mouseover looked like a bug
to someone who was just fooling around with the UI for 15 minutes
without reading much user docs or the functional spec. :-) Sorry, I
may have dragged this discussion for too long. I'd be happy to try out
various designs since this is the one time when Chandler can easily
afford to play with alternatives without upsetting an established user
base, and obviously the user testing that you've been conducting
should be even more useful.
Davor, I'm curious to know what you think
of the insignia on the doorway metaphor I proposed in my last
email...
I think it's not a bad metaphor, but it depends on a) people
approaching the door in the first place; b) being familiar with the
metaphor; and c) the door recognizing when someone approaches it. A
user will approach the collection with the intention of selecting it
(a), I suppose, so that when the mouse comes over it the icon changes
(c), and so (b) remains the main challenge. Maybe the mouseover icon
could be bigger and show a square and a checkmark instead of a circle
and a checkmark -- I think would look more "selectable"? Or use an
icon of a pin?
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Open Source Applications Foundation "Design" mailing list
http://lists.osafoundation.org/mailman/listinfo/design
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Open Source Applications Foundation "Design" mailing list
http://lists.osafoundation.org/mailman/listinfo/design
|
