Usability - SIG, Spin, Echo icons [was Re: Icon theme for ubuntu]

[David Nielsen]

ons, 21 11 2007 kl. 22:31 +0100, skrev Martin Sourada:

> I think, David, if you know about issues with Echo usability why not
> point them out? I think we should take a look at it too a reach some
> consensus between our desires and usability, it is completely useless to
> make a icon theme which someone, e.g. colourblind people, will have
> problems using it. I encourage everyone to use same hue for borders so
> that the shapes would be easily distinguishable even if only used
> without colours (i.e. grey), for one. It is probably not enough, but it
> certainly makes them more usable.

I tried that in the past, with less than succesful outcome. Colorblind
issues however should likely be solved by the compiz plugins, that is
free for you guys and solves the problem (hopefully) for the 3 major
kinds of colorblindness without having to have 3 additional iconsets
differing only in color scheme. I'm unsure if there'll be issues with
shape and the appearance of the icons (clashing colors e.g.) but as part
of the testing of any theme it might be nice to run a real world desktop
image through VisCheck[1]. Likewise we should figure out a way to
simulate common vision impairments, people with slightly blurred vision,
basically people who need glasses but don't wear them. I'm unsure of any
research or products being available to aid in this kind of testing - I
suspect we could apply a slight blur filter in gimp to get an idea of
how washed out the desktop and applications look.

Echo has no defined shape, in my limited testing people with poor
eyesight seem to have great difficulty reading information from icons
that lack a strong outline. For those users, and there are a lot sadly,
the kind of look Tango has seems to yield better results. I'll try to
get more test subjects to play with my laptop with the different
iconsets to get more input. Basically though, Echo is not designed with
this in mind and I suspect reworking Echo would be a lot of work.

Aside that I have noticed that emblem use is somewhat odd, you seem to
use both + and * to indicate new and add (last time I checked - this
might have changed). I filed a bug against this and was basically told
there was a reason but not given said reason. I'd love to hear why this
kind of confusion is preferable to selecting a standard akin to what
Tango has done.

> As for the suggested spin, I think it is a good idea and I see reasons
> for it. Yes, surely all the usability features should be available in
> the classical spins as well, but we don't need to enable all of them by
> default. Especially the artwork could be optimised specially for people
> with disabilities and enabled on such a spin. BTW. is there a SIG for
> usability? As you say, we (Fedora Project) care about usability but the
> applications does not support it as we'd like, so it would be good to
> coordinate our efforts to improve the situation.

I fundamentally disagree with making handicapped people second class
citizens. It's perfectly simple to opt-out, at-spi doesn't draw many
ressources and we need it for section 508 compliance. It makes more
sense to enable it by default and make the cases like colorblindness a
simple configuration option in the a11y capplet that would enable the
compiz plugin and configure it for the sight issue. 

Similar with more extreme sight requirements like high contrast, we can
provide it in the theme capplet. The main problem here is that not all
of our applications obey the session settings, not to mention there's no
option to make it apply to the system. For a single user system, being
able to read the test and see the icons even at boot up would be the
desirable long term goal. I'd rather work towards fixing those problems
and selecting a good default - making a specific spin just for every
combination of a11y friendly artwork seems excessive, for special cases
it's definitely a good option though, like special tool requirements.

I am unaware of the existance of an a11y SIG but usability was handled
on the desktop-devel list way back in the day and Daniel Durand I think
lead up a SIG on the subject. I think ultimately though that the board
(or who ever makes technical decisions like this, there are so many
acronyms and groups I get confused as to who does what) should make the
decision if accessibility is an overall longterm goal to be prioritised
in Fedora or not - then the SIGs can take it into consideration. I will
naturally be happy to help with testing. Given time to schedule testing
it's likely I can arrange to get any such testing done on demand, I'm
fortunate enough to have good relations with a company who exclusively
hire handicapped people so we should be well covered in terms of
impairments. The access is likely going to be timelimited though, these
people suffer and I'd like to not put them through to much so we
definitely need to setup some kind of test protocol - this however is
not really an artwork issue, we can test most poor sight via
simulation.  

[1] http://www.vischeck.com/vischeck/ 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: Dette er en digitalt underskrevet brevdel
URL: <http://listman.redhat.com/archives/fedora-art-list/attachments/20071121/faf2b48d/attachment.sig>