Re: [Usability]Re: Toolbar editor

[Arnaud Charlet <charlet ACT-Europe FR>]

> It scales because you're not supposed to "guess" in the first place,
> you're supposed to gather sufficient user requirements that you "know"
> :)

And again, I disagree.
First this kind of gathering can be very costly. By forcing developers to
make such study, you basically kills most of the software developed for fun
by hobbyist, plus lots of software were time and budget are constrained
(which means basically the rest :-)

Then as I said previousely, different users have different needs, either
because some are experienced users and some are not.

You study in particular simply can't handle properly experienced users of
your software, since by your definition, the software won't be released
until the study is done...

Also, simply because users will have very different needs means that trying
to find a common set suitable for everyone is bound to fail in many cases,
again, for complex applications.

So your approach is only viable economically and technically for small
applications, or for only some kinds of software. It is by no means a
generally suitable approach.

> I'm not disputing that a toolbar editor is probably a very useful thing
> to design well and standardize on.  I guess I'm just worried that:
> 
> -  it might turn up in applications that shouldn't need one "just
> because it's cool", thus adding complexity that needn't exist

Why do you really want to make this decision instead of the user ?
This is really the wrong attitude.

Yes, you can make useful guideline based on experience, but simply forcing
people to do what *you* consider to be the right thing in 2003 (and will
likely be considered obsolete and misguided in 2 or 3 years) is the wrong
approach.

> - developers might use it as an excuse not to design their toolbars
> properly in the first place because they think the user will fix it to
> suit themselves... whereas past experience would indicate that they
> mostly won't bother, so they'll just be left with a poorly-designed app
> on their hands instead.

Fine, you can't stop people from making bad design decisions in the first
place.

Since you are working in a real company writing real software, I'm sure you
should be able to understand that.

For instance keeping your requirements in mind, Sun would never have
developed Java and the Java API, which is clearly a badly designed API,
at least parts of it :-)

Arno