Virtual QWERTY Keyboards to be used with Fingers...
richard5 at email.arizona.edu
Mon Mar 3 19:30:30 CET 2008
To change between character types on the keyboard flick on the keyboard up
or down. To backspace drag across the keyboard back.
From: community-bounces at lists.openmoko.org
[mailto:community-bounces at lists.openmoko.org] On Behalf Of Karsten Ensinger
Sent: Monday, March 03, 2008 5:31 AM
To: List for OpenMoko community discussion
Cc: "Marco Trevisan (Treviño)"
Subject: Re: Virtual QWERTY Keyboards to be used with Fingers...
Marco Trevisan (Treviño) wrote:
> Karsten Ensinger ha scritto:
>> I am afraid, I have to admit that I have NOT tried the Trolltech stuff
>> at all. Maybe because I thought I would deceive the openmoko movement
>> by trying "other" stuff on the Neo. ;-)
>> I will give it a try until next weekend and will change my mind if
>> it is as cool as you claim.
> Ok, we're waiting for your report...!
OK, I installed the latest reviewed u-boot, kernel and rootfs.
Then I tried to install Qtopia on the SD-Card. Unfortunately
every "scp" failed (even if my laptop "thought" it was successful).
I made some test and dicovered that it was possible to copy
arbitrary stuff from flash to SD-Card but not via USB->scp.
I tried to format the SD-Card, but after successful (means: no
error messages) formatting, it was impossible to mount the SD-Card.
Even my laptop (Ubuntu based) denied to mount the SD-Card then.
Lucky me, a new build appeared at buildhost and I tried the latest
and greatest u-boot, kernel and rootfs.
SD-Card could be formatted without any problem and even "scp"
succeeded this time. I followed the instructions on the wiki and
configured the Qtopia stuff for manual starting.
Then I started Qtopia (without a SIM card inserted) and had a look
at the notes application (this was the first application I found
which offers text input).
The keyboard contained only lower case characters and there was no
obvious way to get upper case letters. I was also missing a
possibility to remove mistyped characters.
So the first impression was: an "intuitive user interface" looks
different to me.
Although the character distance was very small, I managed to get
a "hit rate" of more than 90 percent for hitting the intended
The word prediction was great when writing simple english stuff,
but absolutely useless, when typing german. In addition to this,
it missed most of the complicated english vocabulary I tried.
I could not find a way to switch a dictionary, so this was not
Survey (after half an hour of testing and three complete screen
I don't want to put down the implementation of the Qtopia
keyboard at all. I have much respect for the quality one can
But to me it seems as if the keyboard was not designed for finger
use in first place. A "hit rate" of 90 percent (when using an adult
fingertip) is barely acceptable.
The usage is not yet intuitive (how to do a backspace, a delete,
use upper case letters, use special characters (like @,$,&,/)?)
and due to the "look alike" of a regular "hardware" keyboard and
the expectations caused by this, the frustration is high, when
one can not handle simple things without looking for help
(is there a helpfile for the keyboard input? I only found help
for the application itself but not for the usage of the keyboard).
The GUI of Qtopia itself looks very impressive from a design
point of view. But I am missing something like a bubble-help
(e.g. press and hold a key and get a small hint of what the
meaning of the key is), although this is not specific to Qtopia
but is missing in Openmoko also. It is NOT always true that
a picture says more than a thousand words (at least not pictures
of 64x64 pixels).
Maybe I am too biased, but most of the critics we discussed
months ago became manifest in this implementation of a screen
keyboard. It is smart, but needs "tricks" to handle the problems
of imprecise finger touches, lack of screen space for great numbers
of keys and fault tolerance. This leads to a learning curve one
has to master before being able to use the keyboard as such.
If the user has to master a learning curve anyway, why not take a
completely different approach, which is designed exactly for the
problems first and foremost? If the approach is different enough,
the user will accept/expect a learning curve (and will tolerate it,
if it is not too steep).
A car is not the adequate vehicle when travelling on a river, even
though it is possible to adapt one.
OpenMoko community mailing list
community at lists.openmoko.org
More information about the community