GIMPsters,

I was away for a couple of days and I see that in the mean more task were volunteered that have an UI impact:

* metadata stuff (jpeg dialog) * iWarp tool (right-on Tor!)
* jitter and smudge

and something about text (layers), that is not so trivial in the UI btw, text, svg and other vector stuff need their own stacking order within a layer and if you want to avoid the question "are they above or in the layer pixels", then they need their own layer type.

anyway, Sven also asked me to come up with UI items for the roadmap:

2.6

* make full use of cairo for the select/crop tools. out go the inverted lines, in come lightened/darkened planes; * solve user request #1, part 1: one menu bar, keep window with menu bar open when no image is open, true floating inspectors, transparency where inspectors overlap the image (found a way how it can be faked);

(beyond?) 2.6

* solve user request #1, part 2: introduce one(!) single-window mode that users can opt for, it has to be done. dock and tear off inspectors + toolbox everywhere (single + multiple window modes); * color-neutral toolbox icons that are quick to recognise and work with, this task is 50% interaction design, 50% graphic design;

really beyond 2.6

* geometry tools merge, this easy-move/rotate/shear/etc. then also goes into the selection-rect tools;
* unleash the power of GEGL, full access to undo/redo any manipulation * layer dialog renovation;
* all plugins to heads-up displays (no dialog) with full image preview (where performance permits); * blobs of paint palette

and there is lots and lots more to do...

--ps

founder + principal interaction architect man + machine interface works

http://mmiworks.net/blog : on interaction architecture

Sven wrote:

* solve user request #1, part 1: one menu bar, keep window with menu bar open when no image is open, true floating inspectors, transparency where inspectors overlap the image (found a way how it can be faked);

I don't think that "true floating inspectors" are implementable at all.
But perhaps you need to explain first what "true floating" means.

let's see: always on top of any normal window, but under menus and dialogs; does not appear in taskbar; not visible when GIMP is not the foreground application. that is the mandatory bit, I think.

I wouldn't like to put something on the roadmap that we can't solve for
technical reasons.

I was getting more positive because you were talking in optimistic terms about when 'always a window with menu bar' would be implemented. are we now back to 'you'll have to patch all window managers to do that'?

Transparency is not a problem on the platforms where it is supported and
we can easily enable this in trunk right away since we are using GTK+ 2.12 there.

I get the feeling that all is not rosy there, but at least I know that I can always work around that...

--ps

founder + principal interaction architect man + machine interface works

http://mmiworks.net/blog : on interaction architecture

Sven Neumann wrote:

Alexandre Prokoudine wrote:

I'm not sure this is the right time and place to argue, but from my experience using Paint.Net, which has this functionality, transparent palettes will absolutely useless and even annoying.

I agree. Transparency will be distracting.

counter example: Aperture. it is all about using the right gray and alpha values.

The really fat bonus of implementing this is that users get the impression (and associated speed-up in workflow) of being able to see the image on the whole screen. Even if part of the real precision work can only be performed in uncovered areas like now.

Also, because the toolbox and inspectors are still visible, there is no penalty in how quick a click in either can be performed (this is the case for hiding minimising toolbox and inspectors).

As it causes wrong color
appearance, it is IMO a very bad thing to do in a user interface that is
made for working with colors.

well, obviously where one works with colors in the inspector that part is either never or on mouse-over not going to be transparent.

--ps

founder + principal interaction architect man + machine interface works

http://mmiworks.net/blog : on interaction architecture

Sven Neumann wrote:

well, obviously where one works with colors in the inspector that part is either never or on mouse-over not going to be transparent.

Where we are back at the point where this is not likely going to be ever
possible on all supported platforms using the toolkit that we use and will continue to use.

I am used to development teams telling me 'can't do that', but I am not used to them refusing to sit down with me to find a work around solution.

It would probably help if your team seeked for less drastic changes. That would have the benefit that they might really be implemented one day. I really don't want to show users mockups of stuff that they will never receive.

I tend to get 90% results in the projects I work on by setting 100% goals and working with the developers after they tell me we can only achieve 40%. That's how it goes for the last 10 years...

--ps

founder + principal interaction architect man + machine interface works

http://mmiworks.net/blog : on interaction architecture

Sven wrote:

On Thu, 2007-11-08 at 00:06 +0100, peter sikking wrote:

Where we are back at the point where this is not likely going to be ever
possible on all supported platforms using the toolkit that we use and
will continue to use.

I am used to development teams telling me 'can't do that', but I am not used to them refusing to sit down with me to find a work around solution.

We aren't refusing to find a work around solution. But so far we only talked about the short-term plans and we have not sat down at all to talk about the long-term goals for the user interface. Perhaps we should
do that before we put these things on our roadmap.

So for now it is probably best if we concentrate on the short-term goals
that we agreed on and that can be implemented. Perhaps we can discuss the long-term GIMP UI vision at the next LGM then.

fair enough. I am also uncomfortable with the fact of having to convince people of putting on the roadmap rather deep-going paradigm changing solutions without writing first a couple of white-paper type of blog entries where I show how all this stuff hangs together in a deep, complicated way.

--ps

founder + principal interaction architect man + machine interface works

http://mmiworks.net/blog : on interaction architecture