Request for usability assistance

[Jacob R Rideout]

Another question,

Expected behavior: if you choose "ignore all" should that correspond
to ignoring the word...

1A) in the active document (for the remainder of that programs life)
1B) in the active document (for all time)
     [may not be feasible to implement fully at the libs level]
2A) In all open documents of the program for the programs life
2B In all documents that program ever uses
3) System wide, permanently

What other information/description/cues would make it better for the
user to understand.

There usually is an "add to dictionary" option as well. (This would be 3)

For the Ignore all:

KSpell2 uses 2B.

The last time I used MS Word it used 1A, OO.org does too.

Thanks, Jacob

On 12/25/06, Jacob R Rideout <kde at jacobrideout.net> wrote:
> Hello,
>
> I am refactoring and enhancing KSpell2 for KDE4. It will now be part
> of a larger linguistic framework called Sonnet.
>
> This is a great time to address the GUI components.
>
> There are two common ways to handle spell checking:
>
> 1) inline, with some visual indication of underlining. Context menus
> supply suggestions or other relevant data and actions.
>
> 2) A dialog driven search. This dialog usually show the context of the
> word, has suggestions and offers a way to manually override the text.
>
> Also, there is usually some way to add unknown words to a local
> dictionary. Additionally, we are adding a grammar checking component.
> What is the best way to handle the overlap in the interfaces?
>
> So, is there a better way? What can be improved with the current
> interface? Spellchecking is usually designed with indo-european
> languages and scripts in mind, what could be improved for other
> languages?
>
> Suggestions of any kind are welcome.
>
> Much thanks,
>
> Jacob R Rideout
>