Comment 1 by jteh on 2011-02-02 20:59
I've really no idea how to fix this satisfactorily, as I don't have a Handy Tech display.

For the keys which should work but don't (e.g. b2+b4+b5), what does NVDA say when you press them in input help?
Changes:
Milestone changed from None to near-term

Comment 2 by Bernd (in reply to comment 1) on 2011-02-02 21:06
Replying to jteh:

I've really no idea how to fix this satisfactorily, as I don't have a Handy Tech display.

hmm, may you could contakt Felix Grützmacher from Handy Tech to solve this.

For the keys which should work but don't (e.g. b2+b4+b5), what does NVDA say when you press them in input help?

NVDA sais only: b4+b5+b2

If I can help you somehow with testing and providing logs I will do what I can.

Comment 3 by jteh (in reply to comment 2) on 2011-02-02 21:15
Replying to Bernd:

For the keys which should work but don't (e.g. b2+b4+b5), what does NVDA say when you press them in input help?

NVDA sais only: b4+b5+b2

Odd. This should work. Can you please provide the log entry for the key when pressed in input help with log level set to input/output?

Comment 4 by Bernd on 2011-02-02 21:28
ok, here is the log entry. I pressed esc+enter for reference but I can't see any different between the key presses. Esc+Enter works but the b2+b4+b5 doesn't work.
IO - inputCore.InputManager.executeGesture (22:22:44):
Input: br(handytech):b4+b5+b2
IO - inputCore.InputManager.executeGesture (22:23:01):
Input: br(handytech):esc+enter

Comment 5 by jteh on 2011-02-02 21:43
Okay. Found the bug.

Technical: inputCore.normalizeGestureIdentifier assumes that the last key in a combination should be last, even though the order of modifiers is indeterminate. However, for all current braille display drivers, all keys are equal; there are no separate modifiers, so all keys have indeterminate order. inputCore.normalizeGestureIdentifier somehow needs to be fixed to handle this. I'm damned if I know how just yet.
Changes:
Milestone changed from near-term to 2011.1

Comment 6 by bramd on 2011-02-02 23:15
Could you send me a log on input/output level when you press space? And what would be a logical function to assign to it?

Comment 7 by jteh on 2011-02-03 07:47
Ok, the bug with b4+b5+b2, etc. should be fixed in 12f16ac.

Comment 8 by jteh on 2011-02-04 21:08
@bernd: We need a response on this in the next few days if it is to make it into 2011.1, as we are about to enter a translatable string freeze. To recap:

• I don't think we should unbind leftSpace and rightSpace just because they aren't present on one display.
• That said, is there a binding for upArrow key and downArrow key that would work for everyone?
• What do you think is an appropriate binding for the space key on your display?
  • If you want it bound, what does NVDA report in the IO log when you press the single space key?

Comment 9 by Bernd (in reply to comment 8) on 2011-02-07 18:07
Replying to jteh:

@bernd: We need a response on this in the next few days if it is to make it into 2011.1, as we are about to enter a translatable string freeze. To recap:

• I don't think we should unbind leftSpace and rightSpace just because they aren't present on one display.

that's right.

• That said, is there a binding for upArrow key and downArrow key that would work for everyone?

hmm, you could use all other bindings without spaceleft or spaceright.

• What do you think is an appropriate binding for the space key on your display?

hmm, I don't have an idea at the moment.

• If you want it bound, what does NVDA report in the IO log when you press the single space key?

IO - inputCore.InputManager.executeGesture (19:05:04):
Input: br(handytech):space

Comment 10 by jteh on 2011-02-08 09:32
Deferring to 2011.2, since we won't have a definite response on what needs to be changed before the freeze.
Changes:
Milestone changed from 2011.1 to 2011.2

Comment 11 by jteh on 2011-04-20 01:59
Closing as fixed, as the major problem (keys that should work but don't) is fixed. If you want additional bindings, file a new ticket including suggested bindings.
Changes:
Milestone changed from 2011.2 to 2011.1
State: closed