Le 2011-10-06 à 16:35:00, Jonathan Wilkes a écrit :
How do I make a connection from [route]'s rejection outlet to [s]? With the current DesireData keyboard system I select [route] and do <ctrl-tab>9 times, then <ctrl-down-array>, then <ctrl-tab> 2 times, then Enter.
Well, I didn't really work that much on the keyboard-only interface... so I quickly forgot this part. Much of what I did was provide a half-working framework named DesireData to get that project done, and also I coded the part that does find-the-closest-object-in-that-general-direction.
This is the kind of design that benefits a lot from trying it, so, when someone develops such a thing, there is a big advantage to already have the framework for it before starting, and the designing the thing step by step while trying it on real example patches. That way you'll figure out every incentive/disincentive to choose one key instead of another one, etc.
Don't forget keyboard layout issues. Lots of people use a US layout, but people around me tend to use one of the variants of CF layout, which has a bunch of different keys, whereas overseas, the DE layout (QWERTZ Germany) is more different, and the FR layout is even more different (AZERTY where the digits all need shift). At least all of those layouts have the same positions of HJKL (and, of course, of arrows).
Rather than physically clicking a key nine times, I'd like a way to specify the xlet number with the number keys, which could cut the number of keystrokes in half. (That's what I'm talking about regarding vi efficiency, though it doesn't have to be the exact same key bindings.)
You could also have wrap-around topology, using a keyboard shortcut for going in the backwards direction. When you try to go left of the leftmost inlet, you appear to the right.
Wrap-around topology : http://www.gamocratie.fr/wp-content/uploads/2009/10/Pac-Man.jpg
| Mathieu BOUCHARD ----- téléphone : +1.514.383.3801 ----- Montréal, QC