Re: UI Guidelines -- What I'm doing

[Seth Nickell <snickell stanford edu>]

> Of course. We are also, of course, in a somewhat different position than the
> authors of the MacOS guidelines: We have a large base of software already
> written, and at least part of the UI guidelines work will probably be to

This is not quite true. The Macintosh HI guidelines do change from OS
revision to OS revision, sometimes in very significant ways (MacOS8
applications are really different from MacOS6 apps!). Every time they do
that they need to weigh the cost of changing the existing base of
software against the desirability of the improvement. In fact, their
position writing a guide today is similar to ours *except* their
existing codebase is already somewhat conformant to a previous
specification, insomuch as any deviance from the norm will force many
applications to change in that regard.

> > I think we should go into this knowing the guidelines will need to be
> > completely rewritten post-1.0. Having some guidelines is better than
> > none, and will improve the state of UI affairs (hopefully) for GNOME 2.x
> > releases (probably not 2.0, but hopefully revisions). During that time
> > we will learn what we did wrong and revise, and hopefully GNOME 3.0 will
> > be based on much better guidelines.
> 
> I think completely rewritten is being very pessimistic. We're standing on
> the shoulders of giants in many ways: We have the Apple Human Interface
> Guidelines, as well as other documents, to support ourselves when we do

I am confident that the final product would be *MUCH* better than what
we will write as our first "in use" revision if we simply took the
MacOS8 HI guidelines and tweaked them for GNOME. They will make fewer
errors, and will result in better UI. The reason for this is when we
write our own guide, unless we intend to copy the Macintosh guide, we do
not stand on the sholders of giants. The reason is this: pretty much
none of us have completely assimilated the concepts, theories, or even
the practical examples of those that have gone before us. If we were
formally experienced in HCI we would indeed be standing  on the
shoulders of giants. As it is we are trying to shimmie up the giant's
leg ;-) So I expect our own UI guidelines to be extremely flawed because
we do not leverage this experience fully, and hence probably require a
near total rewrite.

HOWEVER! I am *not* saying we should just rewrite the Macintosh
guidelines, even though I think the end result would be better. And the
reason is that we will gain valuable knowledge and experience that will
allow us to do better the next time. And better still the next time.
Until (maybe in the future) we are writing guidelines that are as good
as anything the Macintosh team produces. We need to excercise our UI
muscles if we ever hope to have them be competitive. We're still
learning how to do that with code, and we're a lot more experience with
code than we are with UI right now, so we have a long way to go. Expect
to throw one away, its a common principal of software engineering even
when you are starting with a lot of experience (which we aren't).

> rewritten, etc. That's the way to sloppiness.

That's the way to avoid rigidity.

cheers,

-Seth (grumpy and sleep deprived ;-)