Difference between revisions of "83Plus:BCALLs:50E9"
From WikiTI
(This is about the TI-Keyboard, folks) |
m (categorize this under Link also) |
||
Line 1: | Line 1: | ||
− | [[Category:83Plus:BCALLs:By Name:Input|KeyboardGetKey]] [[Category:83Plus:BCALLs:By Name|KeyboardGetKey]] [[Category:83Plus:BCALLs:By Address|50E9 - KeyboardGetKey]] | + | [[Category:83Plus:BCALLs:By Name:Input|KeyboardGetKey]] [[Category:83Plus:BCALLs:By Name:Link|KeyboardGetKey]] [[Category:83Plus:BCALLs:By Name|KeyboardGetKey]] [[Category:83Plus:BCALLs:By Address|50E9 - KeyboardGetKey]] |
== Synopsis == | == Synopsis == | ||
'''Unofficial Name:''' KeyboardGetKey | '''Unofficial Name:''' KeyboardGetKey |
Latest revision as of 14:16, 6 November 2006
Synopsis
Unofficial Name: KeyboardGetKey
BCALL Address: 50E9
Minimum OS Version: 1.15
Attempt to receive a keypress from the TI-Keyboard.
Inputs
- None
Outputs
- A contains status code
- The actual data retrieved from the keyboard appears to be trashed in the process of popping the error handler ?!
Destroys
- AF, BC, DE, HL
Comments
This B_CALL was introduced in OS 1.15. Programs should check the OS version before attempting to use this routine.
Keep in mind that the keyboard communicates using a truly horrible perversion of the DBUS protocol, which works as follows:
- The keyboard sends the byte E0.
- It pulls both link lines low simultaneously for a short time (i.e., flags a DBUS error.)
- It sends the byte 01h.
- Finally, it sends the scancode or modifier mask.
As a result, there are a variety of possible error conditions.
Possible status codes returned are:
- 00 = No link activity detected.
- 01 = Success.
- 02 = Link activity detected, but the first byte was not E0h, or E0h was not followed by a DBUS error.
- 0F9h = Link assist error, and no buffered data or link activity.
- 0FAh = Link assist error and buffered data detected, but the buffered byte was not E0h.
- 0FBh = Link assist error and buffered byte E0h detected, and two more bytes might or might not have been read.
- 0FEh = Second byte was not 01h (possibly unsupported future commands)
- 0FFh = Protocol error other than the one expected.
Links
A full disassembly of this entry point on OS 2.41 is here.