![]() | SUMit Roster Software > Nut's > October 2007 > Navigation | ![]() |
Example:
mp jbr
gives
Employees > John Brown
step | example | ||||||||||||||||||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
The user types an abbreviation. | |||||||||||||||||||||||||||||||||||||||||
The computer finds the best match for the typed abbreviation.
A match is perfect if all characters of the abbreviation are part of a choice and in the right sequence as well.
|
|
![]() |
Leonard Verhoef has inspired me with his thesis
thesis, why designers can't understand their users
and his Dutch books on design theory.
Chapter 26 of his Dutch books,
"Indelen van informatie en navigatie"
describes exactly
which navigation is easy for people..
Verhoef prefers typed commands. It took me almost a year of puzzling to transfer his advice into a design. Verhoef's books are a time bombs. They are haikus, condensed design mantras, who sometimes take months until they dawn upon me. |
![]() |
With Jef Raskin's GOM method it is possible to compute exactly
how much faster it is to type rather than to point.
See chapter "4-2-1 Interface Timings" in the book The Humane Interface |
death of the desktop |
Asa Raskin that commands are very efficient in
a practical demo of Enso, navigation with language.
A disadvantage of Enso is that the user must make up commands and memorise them. |
Dasher | There is no need to make up and remember commands.
In a demonstration of Dasher,
select words by navigating through a language tree
the computer works harder than the user.
Selecting the first couple of characters often suffices in getting at the right word.
The computer shows the list of valid words. The user just needs to navigate through them. I like that: Recognition in stead of recollection. It takes away the need to make up and remember commands. Dasher is very sensitive to errors. Select one wrong character and the computer heads to a wrong trail. |
Levenshtein |
Andrew Hedges
has a
Javascript version of the Levenshtein algorithm.
This algorithm computes the difference between two strings.
Andrew's demo inspires me to the formula deviation = Levenshtein - length differenceIt takes a pencil and several pages of computations to test my formula. The test results are encouraging. This should work, start programming! |