On Sun, 2005-07-03 at 20:17 +0200, Jesper Juhl wrote:
> Ok, just to show you people what I've done so far.
> This doesn't work yet, but it should be resonably close (at least it
> builds ;)

On top of what you sent at first this morning (not the most recent one)
I did some stuff (in the attached patch).  It implements the last bit of
initialization that your earlier one didn't do, but I see you've done in
the most recent one.

Anyway, I get 10 reads out of it, 1 second apart, and it appears to be
getting real data:

10 seconds while tilting my thinkpad around:

x_accel: 372
y_accel: 339
x_accel: 475
y_accel: 396
x_accel: 429
y_accel: 414
x_accel: 441
y_accel: 519
x_accel: 431
y_accel: 525
x_accel: 387
y_accel: 311
x_accel: 389
y_accel: 246
x_accel: 385
y_accel: 266

10 seconds of stationary reads:

x_accel: 368
y_accel: 390
x_accel: 368
y_accel: 391
x_accel: 368
y_accel: 390
x_accel: 369
y_accel: 390
x_accel: 369
y_accel: 391
x_accel: 369
y_accel: 390
x_accel: 370
y_accel: 390
x_accel: 369
y_accel: 392
x_accel: 369
y_accel: 392
x_accel: 368
y_accel: 390
x_accel: 368
y_accel: 391


-- Dave