I love good dictionaries, and I am hopping between linux and windows all the time, so this is a dream come true for me.
May I suggest something:
1. Scan popup
Scan bubble would pop up faster if fewer dictionaries are used than during normal look-up. This is especially true if web dictionaries are included. One normally does not the full set of dictionaries in the pop-up bubble. Two possible ways to implement this:
- use only the topmost dictionary for scan pop-ups; this would not clutter the user settings interface at all
- add a check-box "use for pop-ups" for every dictionary
- foresee a special "pop-up" group to populate it with dictionaries to be used for pop-ups
Having done better than Lingvo in general, you surely can do better with the tutor as well. It would be nice to keep track of look-ups so that user is notified when he looks up words repeatedly. That same user look-up database could be used for training purposes by popping up the words in accordance with weighted probability, using the unsuccessful answers as weights. Something like that. If you are interested at all, we all can discuss the details in this forum.
I have not yet pulled the git repository, but I might see in the future if I can contribute. I am afraid that my c++ skills (which were never good) are rusty after years of python and straight c development, but I shall give it a try all the same.
Once again, many thanks for the wonderful software (I know good stuff when I see one).
Cheers,
Misha