SEO

Hi,


while having much fun with programming my 9238743th mini-test-app with
reading sensors, I came upon a few ambiguities. Since there are no
examples to be found for a handful of sensor types, declaring their
value to carry a certain unit is probably mostly an academic (though
important-to-me) issue at this point. I propose the following patch,
just in case I made the right assumptions:

---- snip ----
Index: sensors.h
================================================== =================
RCS file: /cvs/src/sys/sys/sensors.h,v
retrieving revision 1.18
diff -u -p -r1.18 sensors.h
--- sensors.h 3 Jul 2006 21:17:37 -0000 1.18
+++ sensors.h 6 Jul 2006 17:44:52 -0000
@@ -34,17 +34,17 @@ enum sensor_type {
SENSOR_FANRPM, /* fan revolution speed */
SENSOR_VOLTS_DC, /* voltage (muV DC) */
SENSOR_VOLTS_AC, /* voltage (muV AC) */
- SENSOR_OHMS, /* resistance */
- SENSOR_WATTS, /* power */
+ SENSOR_OHMS, /* resistance (muOhm) */
+ SENSOR_WATTS, /* power (muW) */
SENSOR_AMPS, /* current (muA) */
- SENSOR_WATTHOUR, /* power capacity */
- SENSOR_AMPHOUR, /* power capacity */
+ SENSOR_WATTHOUR, /* power capacity (muWh) */
+ SENSOR_AMPHOUR, /* power capacity (muAh) */
SENSOR_INDICATOR, /* boolean indicator */
SENSOR_INTEGER, /* generic integer value */
SENSOR_PERCENT, /* percent */
SENSOR_LUX, /* illuminance (mulx) */
SENSOR_DRIVE, /* disk */
- SENSOR_TIMEDELTA /* system time error (mSec) */
+ SENSOR_TIMEDELTA /* system time error (nSec) */
};

#define SENSOR_DRIVE_EMPTY 1
---- snap ----

A few comments:

* SENSOR_OHMS: muOhms ... I suppose that's not far off of
superconductivity. For consistency reasons, muOhms still make sense.
9223 gigaohm fit in there, I think that's enough...?

* _WATTS and _*HOUR: mu/micro stuff simply makes sense here.

* SENSOR_TIMEDELTA: I think this comment is plain wrong -- I highly
doubt it returns values in milliseconds. If sysctl(8) is doing it right,
then it's nanoseconds. (No timedelta devices in my vicinity to try it.)


Moritz