The lonliness of the Graphical User Interface designer...

[Michael Alan Dorman]

On Fri, 1 Dec 1995, Melissa Silvestre wrote:

> programmers are known for thinking that the programmer in the next
> cubicle is a good person to ask whether a new whatever they're writing
> is easy to use. 

Would anyone mind if I made offensive generalizations about librarians
here?  It seems to be acceptible when made about programmers.

> In my work with our computer center on our campus web, I see this "I the
> programmer think it's easy to use, so it must be" all too often. 

Or maybe you misinterpret the motivation behind such alleged statements. 
Try imagining their position -- speaking as a programmer, I often end up
making ergonomic/design decisions because _no one else will take the
responsibility of having an opinion_.  _Someone_ has to do it, and I 
usually get stuck with it, despite my best attempts to get one of the 
eventual consumers of my interface.

So -- I've read _The Visual Display of Quantitative Information_, I've
read _Tog on Interface_. It shouldn't be a big deal -- I try to do what
seems to make sense. 

_Then_--once I've invested tons of work in fleshing out the proposal--and
only then can I actually get opinions out of people. 

Even then most of the people I've worked with are frighteningly
inarticulate when it comes to describing why they don't like something. 
So I have to try and figure out what "It's just not integrated" means when
you're talking about a picture next to some text.

Of course, by the time I've been able to coax something worth hearing out
of these people (a long arduous process involving you doing everything up
to and including constructing sentences for the individual), I no longer 
have time to make any changes.

This ignores the situation where you get input like, "Oh, sure, that's
exactly what you want." followed three weeks later by, "We can't use that
--- it's missing item X" for the exact same input screen. 

So, have you always given your programmers explicit, concrete designs
before having them produce an interface? 

Once they have, are you sure to provide them with lucid, concise
critiques? 

If not, you yourself are part of the problem.

Mike.
--
Michael Alan Dorman                                   Head of Systems
mdorman at caldmed.med.miami.edu           Louis Calder Memorial Library
(305) 243-5530                     University of Miami Medical School