ASU keyboards, again

["Marco Trevisan (Treviño)"]

Carsten Haitzler (The Rasterman) wrote:
> On Wed, 27 Aug 2008 04:21:10 +0200 "Marco Trevisan (Treviño)" <mail at 3v1n0.net>
> babbled:
>> Well I have to say that it doesn't work always as expected... :/ I'm 
>> comparing the prediction of the illume keyboard with the one of the 
>> qtopia keyboard and using the same dictionary (well, in different 
>> formats but it contains in both cases the same 446674 words).
>> So, using the qtopia keyboard the correction applies always, while with 
>> the illume keyboard the more the word is long and more the 
>> typo-correction fails.
> 
> how are you using the correction? are you just taking he "most like;y" 1 word
> only - it should list all corrections - the one you want should be there -
> without frequency information it may not be the most likely match though. it
> could also be that the with illume the .kbd sets the "fuzz" value - this is the
> search distance for neighbouring keys. this fuzz value may be too low for your
> fingers? i really need to know more. also the algorithm for sorting "best
> match" is commented with FIXME's - it's very simple, so if it gets the
> correction listed (check the full list by pressing the top-left dictionary
> icon) then it's finding it - just not sorting it that well. if it is not there
> - it's likely a fuzz value issue. a user config for multiplying the fuzz value
> might be good (to increase search space).

Well, generally for small words there's a correction list, but it's not 
always complete and often there are words very different from the one 
I'd like to write, but not that one. So maybe it doesn't search in all 
the dictionary. I could I try that?
However my fingers are not so great...
If you want I can send you my dictionaries, so you'll be able to test 
them in a better way.

> and is it really the same dictionary? how are you creating the illume and qpe
> dictionaries?

They're based on an online free dictionary, then while illume uses a 
list of sorted words, qtopia uses the same list converted to dawg using 
qdawgen (I don't studied a lot that format but if I'm not wrong [maybe 
I'm only mixing my ideas about a dictionary file format] it represents 
the words in some threes where the roots are the composed by the letters 
and every node could be both a father and a child. I guess that reduces 
the size of the dictionary [1,9mb vs  6mb] and maybe its analysis).
However they contain the same words!

>> Using smaller dictionaries the things improves, but my dictionary isn't 
>> complete yet :o!
> 
> hmm - if it isnt working well for you i am interesting in fixing it - so i'm
> probing for info. :)

Nice to ear! :)

Another thing I'd like to suggest you is that imho the backspace/space 
right-left/left-right dragging is too long. If you try writing using 
your thumbs you can notice that is hard deleting a word... Imho they 
should be more sensible.

-- 
Treviño's World - Life and Linux
http://www.3v1n0.net/