freenode/#clim - IRC Chatlog
Search
7:50:55
beach
And I keep thinking of gestures on a tablet that would simplify the interaction with a score editor.
7:52:11
beach
Once when I suggested a student project, the student proposed an interface whereby all the music symbols would be displayed, and the user would drag symbols to the desired place on the score.
7:52:52
beach
I then suggested a word-processor interface to the student, whereby each letter would be displayed, and the user would drag each letter to the desired place in the text.
7:54:50
jackdaniel
sure, on tablets too, but also when typing text in alphabet which is not mapped onto the keyboard (i.e a cyrillic on western keyboard) when someone can't write without looking
7:55:15
beach
I think the absolute horrible thing with current interfaces is that, since it is rare that the same symbol is used twice in a row, the user has to alternate between the place where a symbol is entered, and the tool bar where symbols are displayed.
7:55:20
jackdaniel
not the most efficient way but as a fallback it is something which allows getting thing done
7:56:51
beach
Anyway, I'll be a but busy this morning. A physical therapist is coming to the house in order to try to fix my back that I seem to have injured a week or so ago.
7:58:44
jackdaniel
I had a project for my bachelor's degree where we were writing a tool to asses the keyboard layout efficiency (with two default metrics: typing speed and hand effort)
8:10:40
ck_
jackdaniel: that's interesting. How did you deal with the influence of training/familiarization?
8:13:28
jackdaniel
we didn't; it was all simulation where particular fingers were associated with keys and the effort was split between each finger due a course of whole text and speed was a function of time needed between pressing two keys (they might be the same hand or not, that mattered too)
8:14:37
jackdaniel
distance between keys was taken from the keyboard model created before the simulation (so different keyboard shapes were tested too)
8:15:07
jackdaniel
it is worth noting that the guts of the whole project were written in common lisp
8:17:13
jackdaniel
results of simulations were not very spectacular -- basically all except ridiculus designs had similar efficiency (I think that dvorak had negligible adventage for english text, but I might have misremembered)