Editing Scancode (section)

==IBM PC scancodes==
Scancodes in ''Set 1'' were used over the [[XT keyboard interface]]. They are still in use internally on systems with IBM PC lineage such as [[Microsoft Windows|MS Windows]] and [[Linux]] ("raw mode").

Scancodes in ''Set 2'' are used in the protocol used by the [[AT keyboard interface|AT]] and [[PS/2 interface|PS/2]] interfaces.

Scancodes in ''Set 3'' are used in the keyboard protocol for terminals from IBM.

===History===
The [[XT keyboard interface|IBM PC/XT protocol's]] scancodes corresponded directly to key positions when ''scanning'' the original [[IBM Personal Computer keyboard|IBM PC/XT keyboard]]'s [[keyboard matrix|matrix]]. A single bit in the byte sent over the wire distinguishes between a key press (''make'') or key release event (''break'').
Raw scancode bytes were delivered to programs by the BIOS.

The later protocol used by [[AT keyboard interface|AT]] and [[PS/2 interface|PS/2]] interfaces uses the different ''set 2'' of code sequences over the wire.
For the IBM PC/AT to remain backwards-compatible, [[XT keyboard interface|XT]] scancodes were not only presented to programs by the BIOS but there was also a special microcontroller which converted the new Set 2 codes back to the old Set 1 codes for badly written programs that accessed the hardware directly.

However, with this backwards compatibility came added complexity. The [[IBM Enhanced Keyboard]] (Model M) introduced even more additional keys.
Pressing some keys can produce quite long sequences to be sent, especially with modifiers. Therefore, IBM introduced the cleaner "Set 3" set of codes, also with some more capabilities.
Some IBM keyboards are able to switch between different code sets on command from the host. <ref name="nisley59">
Ed Nisley: '''How the PC Keyboard Got Its Bits'''. Circuit Cellar, issue #59 June 1995. Page 46.
</ref>

The Set 3 did however not catch on outside IBM and was used almost only for terminal keyboards.