[linux-elitists] Re: GNOME > you

Jeremy Hankins nowan@nowan.org
Mon Jan 5 08:51:05 PST 2004


Jeff Waugh <jdub@perkypants.org> writes:
> <quote who="Jeremy Hankins">

>> So when the first axiom in your world view seems to be "but users
>> don't care" I find that very, very significant.
>
> I don't think it's fair to describe it as "the first axiom in [my]
> world view" at all. It was a response.

It's true that you never explicitly presented it as such, but it was
certainly a point you kept coming back to, as if it was the ultimate
support for your perspective on UI.  So perhaps "world view" should be
changed to "world view on UI" -- or perhaps I'm misunderstanding.

> The example covered two people's approach to writing a document
> (what), with one having very specific technical requirements as to how
> he'd go about it, while the other did not share similar requirements
> (how). Perhaps she wanted other things, such as the ability to do nice
> background and coloured text or something.

Ok.  So where do formatting decisions fit into the mix?  Sandy has to
worry about that, I don't.  Sandy must express more information to get
equivalent output.  Is that part of the what or the how?

One way to think about it is to simply define what & how as what I care
about and what I don't, respectively.  Then an interface could be graded
on how much of the how it forces the user to deal with.  This is
probably closest to how we usually use the words.  In other words, the
what is content, the how is syntax.  Is the syntax arcane and wordy, or
does it contain just enough to differentiate the content?

A similar analysis could be made of a wm.  My wm is set up to make
window location irrelevant, as it's something I don't care about.  I
have a set of keybindings that allow me to say "I want app Foo", and
that's what I get (whether it's already running or not).  No extraneous
information being conveyed.  A traditional wm uses location to
differentiate between various apps, which means the user is stuck with
the job of translating "app Foo" into "over there" where "there" is a
location on the screen (possibly under other windows).

(Note that while I place myself firmly in the "no mouse" camp, it's the
"data encoded as location" that I really disagree with -- that's why I'm
skeptical about post-hoc attempts to add keyboard control.)

> The point was that because Sandy didn't have *technical* requirements,
> she "didn't care" about those details, and wouldn't easily dive into
> the idea of using LaTeX for her document. It simply doesn't satisfy
> her (most likely subconcious) soft needs, such as interface
> discoverability and so on. She probably doesn't want to read manuals
> as a prerequesite to getting her task done. The other subject had
> specific technical interests, and was far more willing to learn a
> complicated, largely non-discoverable system to get the results he
> wanted.

Yes.  For many people low-to-no startup investment is the highest
priority.  I think that's as misguided with computer interfaces as it is
anywhere else, but I'm not (well, not really) criticizing gnome for
giving those folks what they want.  If they want wonder bread & jello,
give it to them; they're the ones missing out.  Of course, when I have
to stop using an app because it becomes too gnomeified, I do get a bit
annoyed.

-- 
Jeremy Hankins <nowan@nowan.org>
PGP fingerprint: 748F 4D16 538E 75D6 8333  9E10 D212 B5ED 37D0 0A03



More information about the linux-elitists mailing list