... Palmer¹

mailto:doug@charvolant.org http://www.charvolant.org/~doug

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... something.²

Which should teach me something. Halfway through writing this, I discovered Ivan Pascal's documentation.[6] Still, the more the merrier. In the meantime, I've stolen shamelessly from Ivan's documentation.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... reverse-engineering.³

Undocumented configuration files. Undocumented code. Open-source is all very nice, but this represents a pretty high barrier to entry. It's been pointed out before that a non-trivial open-source project without adequate documentation might as well be close-source.[2]

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... released.⁴

The keyboard itself may generate up to six events for each key press. These events are pre-processed and gathered together before being passed onto XKB.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... down.⁵

The business of locking keys such as the Caps Lock key is handled by the compat component -- see section 4.5. It's enough for a type to know that the key is active.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... 47cm.⁶

Yes, the XKB protocol specification says 0.1mm. Looking at actual geometry files, though, reveals this to be rubbish, unless people really use 4.7cm wide keyboards and I'm in a strange dimensional vortex. Since the results scale in Postscript to fit a piece of paper, I suspect that it doesn't matter.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.