OpenMoko - SoC--- is there a mentor?

Karsten Ensinger at
Sun Mar 25 10:01:13 CEST 2007

Hi Guy,

the linchpin of you concept is the keyboard layout. If you use
unsuited layouts initially, no one will invest the time to get it
right for themself, because one can't easily experience the advantages
of the concept when making first impressions.
Although I am not an expert for language forensic, I assume that the
biggest part of your work will NOT the implementing part, but the
research for a propper keyboard layout for english and portuguese at
A lot of fine concepts in open software got never used due to the
circumstance that the first available implementation couldn't impress
the users enough to encourage them to invest their time to improve
As you said on your web page, your concept is a composite of the
ideas your found on the wishlist. If I identify the "original"
ideas correctly, your solution will need more time to get used to
as the single "original" ideas will do. So you have to offer a
significant added value, which would be a working keyboard layout
for at least the english language (I assume that the majority of
users will be native english or correspond in english mostly).
And do think about the implications of different layouts for the user.
For every language one has to learn a completely different position
of each letter. What is left from the advantage of the "single stroke"
architecture if one has to search for the next character each time
after changing the layout to a different language?
We end up at the same point we started: the layout is the linchpin
(and -> in my opinion <- also the weak point) of your concept.


--- gsilva85 wrote:
> Karsten,
> i agree with you about the layout of the chars ( on my example they was 
> put aleatory )! but at this point I'm just introducing another concept 
> of a finger based keyboard. I'm not able to know whats the better place 
> for each char optimized for English, Germany or even to Portuguese (my 
> native language, I'm Brazilian...).
> For this use cases I'll try to implement some way to easy develop/change 
> the layout of keys (like i said at bottom of my page). Anyone could make 
> their own layout optimized for his language, share with community then 
> you download to your phone, when you need it probably you press some 
> predefined key on this keyboard, select what layout you want and 
> dynamically change to it. This concept is already presents in phones, 
> like when you change the input method to letters, number, symbols, t9...
> what do you think?
> Guy
> PS:don't worry I'm still encouraged to submit it to SoC : )
> 2007/3/24, Karsten Ensinger < at 
> < at>>:
>     Hi Guy,
>     following your explanation, I will get "c", "z", "g" and "q" as
>     characters when tapping on "1", dragging to "3" and dragging back
>     to "1". When I want to have "c", "g" and "q", I have to tap on
>     "1", drag to "2", lift the finger, tap on "3" and drag to "1".
>     If I want to have "c", "g" and "c", I have to tap on "1", drag to
>     "2", lift the finger, tap on "3", drag to "2", lift the finger,
>     tap on "1" and drag to "2".
>     Depending on the needed character sequence, your "new" concept
>     can result in the same amount of taps and drags as the original
>     "finger splash" concept.
>     So everthing depends on the intelligent placement of the characters
>     on the keyboard layout.
>     Let's come to the usability aspects.
>     If I look at the "finger splash", the tapping on "1" will result
>     in a button overlay which will hide the complete "2" at least
>     (maybe also some parts of "3", depending on size). So there is no
>     way for the user to see what is on the right side of the "2"
>     button. This leads to a drag into the blind to get the "z".
>     This seems to me as not user friendly enough to get accepted
>     widely. One has to remember the whole keyboard layout in mind
>     while typing.
>     This would lead to the idea to NOT enlarge the buttons when one
>     taps them. The user would be able to see what he gets, when
>     dragging to the next button.
>     Unfortunately the size of the buttons can't get big enough to
>     hold characters with an easy readable fontsize due to limited
>     physical screen size. This seems to be a dilemma. Making the
>     buttons bigger means less buttons per row and column, means
>     less benefit from the "dragging feature".
>     Maybe the KISS pattern matches here (KISS = Keep It Small and
>     Simple). Although the ability to type more than one letter with
>     a single stroke has charming aspects, the learning curve of the
>     keyboard usage should be as steep as possible and to me, the
>     current concept seems to be to much in need of an explanation.
>     What I have only shortly mentioned before, is the fact that you
>     have to analyse the inherent syllables of the language the user
>     will use. You have to place the characters in a way, that one can
>     type as much words with one or two strokes as possible. So the
>     keyboard layout will vary from language to language.
>     What about users using two different languages at the same time?
>     They will have to pay the price for this. One language will
>     fit perfectly to the keyboard while the other will not (most
>     of the times).
>     I myself will definitely use german and english at the same
>     time every day, and I think most of the non-english natives
>     will do the same. German and english do not have a lot of
>     syllables in common when counting the frequency with which
>     they appear in day to day communications vocabulary.
>     But whatever I comment here, it's my personal and therefore
>     biased opinion. I do NOT want to discourage you to start your
>     own implementation, because an old man think he found a fly
>     in the ointment.
>     Take my comments as food for thoughts and nothing else.
>     Regards
>     Karsten
>     --- gsilva.85 wrote:
>      > Karsten,
>      >
>      > you are right except for one thing on my idea: you are
>     considering only
>      > a row and only a drag from left to right (or right to left..) in
>     this
>      > case only 2 characters will be draw. But think if you press the
>     '1' drag
>      > to '3' and drag again to '1' we got one press with four characters at
>      > output but on finger splash to produce 4 characters (in the
>     better case)
>      > we have to press 4 times. And you can drag not only to left or
>     right but
>      > up, down...
>      > did you understood?
>      >
>      > I thought to adapt this to fingers splash: when the 'splash'
>     appear if
>      > you continue dragging over the buttons of the splash it possible
>      > continues  writing that chars (like the speed script concept) but
>     even
>      > with this we stay limited to only 7 different chars in a drag...
>      >
>      > please fell free to comment...
>      >
>      >
>      > thanks
>      > Guy
>      >
>      > [...]

More information about the community mailing list