Re: [Usability] GIMP - CSDI and Toolbox

[Havoc Pennington <hp redhat com>]

On Tue, 2004-06-15 at 08:47, Marek Peteraj wrote:
> One major flaw of the HiG is that it tends to prefer certain models and
> encourage them.

That's not a "flaw" it's the whole point.

We need consistency on the set of overall app archetypes for multiple
reasons:
 - consistency for users
 - so we can implement the desktop

We can't change the WM to handle 2000 different kinds of windows and
sets of window behavior. We need to know what the kinds of apps are,
what kinds of windows those apps have, and how they behave, and app
authors don't get to make up more an an ad-hoc or per-app basis. They
can make up more on a global basis, though.

Now, I'm very open minded about what the kinds of apps and windows are,
*as long as it's written down* in a holistic and logical way.

> 1. i think the HiG should reflect their work, not
> vice-versa(re-thinking, perhaps reinventing the wheel ui-wise or
> "forcing", telling gimp devs or devs of other apps that might take use
> of the gimp CSDI UI model, to do this or that without any or not enough
> knowledge of the target userbase).

I'm all for the HIG reflecting the GIMP's work, as long as it's written
down in a holistic and coherent way.

By holistic/coherent I mean for example that there are clear guidelines
on when to use one archetype or window type and when to use another, and
how those window types are supposed to behave, and each different
archetype or window has a usability (rather than political) reason for
existence.

> 2. As the gimp represents a reference CSDI design(which i think is no
> surprise since it's been 7 years of CSDI Gimp), it's already possible to
> start working on WM behaviour for CSDI.

Only if somebody writes down how it's supposed to work, separating
historical accident from intentional design.

I'd certainly be all for WM patches in this area, but I think a fairly
broad effort is needed to figure out what the patches should do.

Havoc