On 1/8/06, MenTaLguY <mental@...3...> wrote:
Well, it depends on what you mean by "area". The square may actually be larger in the width*height sense, but the longer shape can have a larger "visual footprint".
And easier to click, too. With the stretched swatch, you need to be precise only in one dimension. With a square, you need to be precise (though a bit less precise) in both dimensions.
That's actually been the main thing driving my experimentation here -- I'm finding the narrow strips to be an extremely difficult mouse target. Out of curiousity, bulia, what resolution do you normally run at?
1280x1024, with pen instead of mouse. No problem with aiming at all. However, I would not object to increasing the total height of the statusbar (and maybe running the statusbar tips in two lines, wrapped), with corresponding increase in height of these swatches.
I do think it's better when you don't have to use text to explain what something is -- particularly in such a compact area.
No. This is an example of a very common fallacy. People are taught that "icons are better than words" as if this is the whole truth, even though its applicability is quite limited. Icons are better only when they are few, simple, and obvious, but in many cases you can't achieve some or all of these objectives. Failing to recognize that and insisting on presenting everything "visually" results in UIs overloaded with too complex, cryptic and confusing icons where simple text labels would work much better. You should remember that letters of the alphabet are _images too_, and very special images at that: they have developed for centuries to be immediately and faultlessly recognizable in very small sizes. Few icons can compete with letters for the speed and accuracy of perception.
So, when I see "None" in my indicator, I do not read it; I perceive it as an image, as a whole, and it is _at least_ as fast as the perception of a red diagonal on white. However, compared with the diagonal, "None" has a few advantages: it's more informative for newbies, larger, can never be confused for some exotic gradient, and in general, much more strongly asserts "no fill/stroke" as a separate property that is unlike all other kinds of fill/stroke. Perhaps the most common type of query that I use the indicator for is "whether this object has fill/stroke or not", and from this viewpoint, the difference between "None" and all other fill types (especially color, of course) in my indicator is much stronger and therefore takes less milliseconds to read than the difference between the red diagonal and all other types of stuff that your square swatches can display. The same applies to all other fill/stroke types that I use text labels for, as well as the F/S labels.
I may fall back to using text to indicate gradients, though it's obviously harder to do in a square space like that. Rather than a tooltip showing only the gradient name, though, I think it might be better to include a larger and visually unambiguous representation along with it. It'd require a custom tooltip window, but I think it'd be worth it.
I disagree here as well. I think a large tooltip with an gradient rendition will be very clumsy and awkward.
By the way, my design in principle can show gradients, because there's enough horizontal space for that, and we even have a nice horizontally-stretched widget for that (now used in the gradient tool's toolbar menu). That would be logical, consistent, and much easier to read than in a square swatch. However, I want to display textual labels in that case as well, for the reasons I quoted (such as dark green/light green gradient being hard to tell from flat green color), and the layout of the widget with both gradient display and a label needs some thinking. But I think I'll do it one day.
I would've punted on checkerboards, except I think bulia got away with it in the existing widget just fine. So I don't think they're a problem in themselves.
In existing swatches they are used only for fill/stroke opacities, which are quite rarely used (most of the time, master opacity is much more convenient). In your design, master opacity slider shows checkers _always_ (by the way how will you make it draggable? it looks like it will require a custom widget).
As for the slash, it's not just Adobe, Macromedia's done it forever too. I'd say it's a pretty well-established convention for graphics applications. Admittedly, I don't know about Xara...
Xara has no red slash, but uses just clear checkers in the indicator (i.e. there's no difference between no fill and fully transparent fill). However, for stroke, there's another indicator that shows stroke width, and there it displays "None" for no stroke. That's where I look when I need to find out if an object has a stroke or not.
In short, your variant cramps more information, but it makes it much more difficult to get it, and unless you increase the size of the widget significantly compared to the size it currently has, it will be outright frustrating.
I would almost advocate increasing the size of the existing widget.
And I would not object, see above.
In my experience, that becomes true of anything once you get accustomed to it. Bulia, are you sure this part isn't a "because I'm used to it" thing? :)
Maybe. But I also have other arguments in favor, as you can see :)
I don't plan on implementing anything right now, partly because my ideas aren't fully baked, and partly because I have bigger priorities (keybindings + layer dialog, mainly).
I don't want to sound like I'm trying to talk you off this, but I agree that keybindings and especially the layer dialog are more important, from the viewpoint of Inkscape's general progress.
-- bulia byak Inkscape. Draw Freely. http://www.inkscape.org