OK, so, to avoid million quote-requote, I'll summarize:
- I see kinetic scroll as a specific mouse gesture
- There is no such thing as useless animation. Unfortunately, users I had feedback from seemed to want/like it to whizz woosh and wazoom, even if you just picked an option or pressed a button. Canola is a good example, but fremantle in general also employs large amounts of animation 'for the fun of it'.
- I played with 'blind' kinetic scrolls in pyqtoreader. It's not that hard. It's pretty much equivalent to pressing a couple of keys in advance (like with cursors or pgup/pgdn while the screen has not refreshed yet). The biggest impediment there is that it's simply not cool. It works (especially as a page flip button replacement), but it's too raw.
- maemo 4.x default scrollbars are microscopic and hard to use, agreed, thus making almost ANY scroll solution appear less strenuous
- I might be 'misnaming' drag zoom. A tap or mouse gesture shows the page fit to the screen (with the current viewpoer marked), and dragging the zoomed box positions (scrolls) to the specified point on release.
- hm, two of two with pyqtoreader install problems, I'll check the package, there is a new version in the works anyway. If I had a camera at hand I'd record a video It would demonstrate much more clearly what I'm saying.
About your pyqtoreader, unfortunately I'm not able to install it (some issue with python-qt4-* package versions that apt-get doesn't like), so I cannot comment on the scrolling methods you have implemented.
Hm, installed/uninstalled fine for me... Can you tell me the exact error/dependency problem ?
~# apt-get install pyqtoreader
...
Los siguientes paquetes tienen dependencias incumplidas:
pyqtoreader: Depende: python2.5-qt4-gui pero no va a instalarse
E: Paquetes rotos
(translation:
The following packages have unmet dependencies:
pyqtoreader: Depends on python... but won't be installed
E: broken packages)
~# apt-get install python2.5-qt4-gui python2.5-qt4-core python2.5-qt4-common
Los siguientes paquetes tienen dependencias incumplidas:
python2.5-qt4-core: Depende: python2.5-qt4-common (= 4.4.3-maemo9) pero 4.4.3-maemo7 va a ser instalado
python2.5-qt4-gui: Depende: python2.5-qt4-core (= 4.4.3-maemo7) pero 4.4.3-maemo9 va a ser instalado
E: Paquetes rotos
- There is no such thing as useless animation. Unfortunately, users I had feedback from seemed to want/like it to whizz woosh and wazoom, even if you just picked an option or pressed a button. Canola is a good example, but fremantle in general also employs large amounts of animation 'for the fun of it'.
What would be an example of such large amounts of animation "for the fun of it"? Most animations used in GUI design fall into the category of transitions. Transitions generally help the eye to follow what's going on and make interfaces more pleasant to use (which is what usability is all about). Transitions can be more elaborate than they have to, which is often done purely for effect. But I doubt you will find much of this in Fremantle.
Originally Posted by
- I played with 'blind' kinetic scrolls in pyqtoreader. It's not that hard. It's pretty much equivalent to pressing a couple of keys in advance (like with cursors or pgup/pgdn while the screen has not refreshed yet). The biggest impediment there is that it's simply not cool. It works (especially as a page flip button replacement), but it's too raw.
It may not be hard for you, but blind mouse gestures are already a bit hard to discover. Making the amount of scrolling dependend on how fast you execute the gesture without any visual feedback... that's intense. Of course if you are designing for geeks, that can certainly be described as "not that hard".
Originally Posted by
maemo 4.x default scrollbars are microscopic and hard to use, agreed, thus making almost ANY scroll solution appear less strenuous
Of course large scrollbars help the issue. Content panning scores points here by offering an even larger drag target, while reserving the largest amount of space for content.
Originally Posted by
I might be 'misnaming' drag zoom. A tap or mouse gesture shows the page fit to the screen (with the current viewpoer marked), and dragging the zoomed box positions (scrolls) to the specified point on release.
That's a good one, but would you want this as a universal replacement for kinetic panning? It's a two-step process, which is much less efficient if you just want to slightly adjust the view. And it's usefullness depends on the size of the content. If the content is only slightly larger than the view, or infinitely large, then it's not so effective.
Kinetic panning makes for a good standard, which can certainly be augmented with more specialized forms of scrolling if the type of content allows it.
- I see kinetic scroll as a specific mouse gesture
You are right, from a programing point of view.
From a ergonomic POV, you have three different pointing devices (finger, stylus, and mouse), each one used in its own way and with its own issues. So, kinetic panning is very desirable when using the first two, specially on hand sized screens because the angle of the movement required. But, with mouse devices, it isn't so desirable. So, I see it as a "hand/stylus gesture"
What would be an example of such large amounts of animation "for the fun of it"? Most animations used in GUI design fall into the category of transitions. Transitions generally help the eye to follow what's going on and make interfaces more pleasant to use (which is what usability is all about). Transitions can be more elaborate than they have to, which is often done purely for effect. But I doubt you will find much of this in Fremantle.
The excess of animations can be even worse than the lack of any animations. While many animations fall into the category of "useful" (being pleasant could be considered useful too), sometimes they are just distracting or even cause motion sickening (I can't stand many compiz effects, and even know people unable to play FPS games due to motion sickening).
It may not be hard for you, but blind mouse gestures are already a bit hard to discover. Making the amount of scrolling dependend on how fast you execute the gesture without any visual feedback... that's intense. Of course if you are designing for geeks, that can certainly be described as "not that hard".
Not only that, but when the page is moving after the gesture, if you see the content passing you can stop the movement, just as you expect by putting the finger or stylus on the screen.
The biggest impediment there is that it's simply not cool. It works (especially as a page flip button replacement), but it's too raw.
You feel it "raw" because your brain expect a certain response or feedback from the "object" after "pushing" it, and is getting a completely different one.
That's a good one, but would you want this as a universal replacement for kinetic panning? It's a two-step process, which is much less efficient if you just want to slightly adjust the view. And it's usefullness depends on the size of the content. If the content is only slightly larger than the view, or infinitely large, then it's not so effective.
Completely agree.
I know this method from "The gimp", where it is used for quick navigation and positioning in a image bigger than the window.
I think it is a good complement to kinetic panning, usable for more or less fast/rough positioning in big pages.
Originally Posted by
Kinetic panning makes for a good standard, which can certainly be augmented with more specialized forms of scrolling if the type of content allows it.
It would demonstrate much more clearly what I'm saying.