View previous topic :: View next topic |
Author |
Message |
johnsfine Site Admin
Joined: 10 Aug 2003 Posts: 4766 Location: Bedford, MA |
Posted: Mon Feb 28, 2005 9:33 am Post subject: |
|
|
jon_armstrong wrote: | I will create a protocol.ini entry when it's done. In fact, that will be important since there is a check nibble that I am computing externally and RM can do that nicely. |
That's the keyboard one, right? And it assumes there is no carry from adding the device nibble to the first keycode nibble (which is reasonable because that "device" nibble seems to be zero for all keyboard samples).
I'm pretty sure the generic check byte logic of RM cannot do the correct wrap around carry. RM is designed to make it easy to add extra Java modules that do things like that. I should write a Java module to do the check nibble calculation with wrap around carry. |
|
Back to top |
|
|
jon_armstrong Expert
Joined: 03 Aug 2003 Posts: 1238 Location: R.I.P. 3/25/2005 |
Posted: Mon Feb 28, 2005 10:00 am Post subject: |
|
|
johnsfine wrote: | That's the keyboard one, right? And it assumes there is no carry from adding the device nibble to the first keycode nibble (which is reasonable because that "device" nibble seems to be zero for all keyboard samples).
I'm pretty sure the generic check byte logic of RM cannot do the correct wrap around carry. RM is designed to make it easy to add extra Java modules that do things like that. I should write a Java module to do the check nibble calculation with wrap around carry. |
I probably wrongly assumed it could I did go back and check all the regular KB keys using the wrap around carry and (as I'm sure you already knew), it is also consistent with them. So I suspect the wrap around carry is the true check nibble logic for all commands but the simpler sum of the nibbles works so far for all keys that have been decoded. So we can start with that until we find a command that won't work.
In thinking about the mouse commands, even though we now could emulate any direction and velocity, as a practical matter with only four arrow keys on OFA remotes (are there any with more?), we probably just want to pick a moderate speed in the four cardinal directions plus the right and left mouse buttons and create an upgrade. Then have users key move those six keys to their regular KB upgrade device. We should be able to do that in a double byte protocol, so we can just calculate those two bytes manually and I'll document the process.
I created an Excel spreadsheet that can calculate the commands and check nibbles for every 15 degrees of bearing at any constant velocity and I'll just add the calculation of the two variable bytes (lsb first.) We could also create the left or right mouse button plus the four direction keys and those could be moved to the shifted (or x-shifted) arrow keys and get a pretty cool emulation of the mouse in order to drag the mouse. Does that sound reasonable? _________________ -Jon |
|
Back to top |
|
|
86turbodsl
Joined: 27 Feb 2005 Posts: 10
|
Posted: Mon Feb 28, 2005 11:51 pm Post subject: |
|
|
I got KM working tonight, and managed to upload a device and protocol up to my remote. It works for the most part, but it still needs some tweaking. The keymoves wouldn't import into IR. I got an error about invalid input when I tried to do a combined keymove and protocol/device with Ctrl V. When I greyed it out, it imported fine. So some buttons dont' work. I think there were 3 keymoves that didn't import that way.
But it's late, and I need to go to bed.
Much thanks for the work that you guys have done on this!!! |
|
Back to top |
|
|
86turbodsl
Joined: 27 Feb 2005 Posts: 10
|
Posted: Tue Mar 01, 2005 8:36 pm Post subject: |
|
|
Ok, I've been playing around with it tonight, and found that the ESC key doesn't work when mapped to the remote exit key, and the spacebar key seems to be missing from the keymap.
Also, keys are not repeating for me as they did for the other person.
Any ideas? |
|
Back to top |
|
|
|