Re: [Usability] CD Ripping Interface [was Re: ui-review of new modules]

[Luis Villa <louie ximian com>]

On Mon, 2003-07-07 at 23:17, David Adam Bordoley wrote:
> Havoc Pennington writes: 
> 
> >> FWIW, the best ui for ripping cd's is probably to put an icon on the
> >> desktop representing the cd, open the cd, dnd the "files" from the
> >> disk to any folder, popup a simple dialog asking the user in what
> >> format they would like the songs to be stored in.  [2]
> > 
> > I don't disagree, but see again this has no replicable rationale or
> > process behind it. How did you arrive at this view; what are the
> > personas, tasks, research, prior art, and so forth. If I wanted to
> > argue with you, or understand your point, or learn from your post, how
> > would I even start. 
> > 
> > Moreover, this view is only in a transient email with a poor subject
> > line; if I were going to implement a CD ripping interface, how would I
> > find it? Instead, it should be somewhere more permanent. 
> > 
> > This is what I mean by creating a replicable process and an output
> > product for usability. An example might be a repository of design
> > documents including rationale and proposed UI. 
> > 
> > (I understand your comment was just an aside, but I'm picking on you
> > to make the point ;-)
> 
> I forgot to reply to this. 
> 
> My design is based on prior art, specifically the copying files from a 
> "floppy disk." Ripping cd tracks from a cd is an identical task to copying 
> files from a floppy disk. Why create a new one? 

[Please continue to not take this personally, Dave; it's certainly not a
fault exclusive to you :)

I think Havoc is more concerned with the meta-problem, Dave- if someone
wants to come in and 'design a GNOME application', how could they do
that? What prior <b>GNOME design</b> art would one look at?

The current gnome usability situation is this:

Luis wants to design a gnome CD player.
Luis has three options:

(1) take classes in human interface design, read the HIG, study all
existing GNOME apps to find similarities, and hope that those
similarities constitute a GNOME design philosophy. Then I apply what was
learned in the classes and design the app.

(2) ask dave b., calum, or seth to design it for him; wait (possibly
eternally) for them to have time and/or interest.

(3) hack up something, have the usability list ridicule him for lack of
design sense/experience.

(1) and (2) don't scale very well, and (3) isn't very constructive.

So... I think what havoc proposed (and I strongly endorse) is the
creation of option (4):

(4) Luis reads the documentation describing what the GNOME design
philosophy/model is. Luis then reads the document describing the
idealized design process for a good GNOME app, which is nicely backed up
by examples on how the usability team or members thereof have designed
other GNOME apps (real or hypothetical) in the past. If Luis has lots of
time, Luis reads some of the other suggested readings he finds on the
GUP page. Luis then designs his app following those guidelines, which is
not perfect (since guidelines cannot completely substitute for
experience) but is a much better approximation of a good app than if
he'd used process (3), and happened much more quickly than (1) or (2).

I realize that (4) would take a lot of time and effort on the part of
the usability team (as did the HIG), but it's the best way (IMHO) for
usability to become part of the GNOME culture.

[FWIW, most of GNOME needs documentation like this- certainly a11y does,
though Brian seems to be addressing that, and better sample apps and
documentation for our core technologies would also be useful. For the
core apps and (to a lesser extent) a11y, though, you can RTFS- that
doesn't help for UI because as of yet there is no S for UI- only
finished products.]

Anyway, my two cents.

Luis