Porting OpenBSD's sysctl hw.sensors framework to FreeBSD

[Poul-Henning Kamp]

In message <20070711003958.V8913 at fledge.watson.org>, Robert Watson writes:
>On Tue, 10 Jul 2007, Poul-Henning Kamp wrote:

>> The OpenBSD stuff is a 1980 style hack, and should not be propagated.
>
>This argument would be more convincing if accompanied by a concrete example, 
>fabricated or otherwise.  Are you suggesting, for example, adding newbus 
>sensor methods associated with existing driver attachments?

I'm not advocating that we actually tro to overengineer a solution for
this stuff.  As long as the hardware people don't think about where
the slam down the sensors, there is little chance of us making any
kind of sense of their measurements without the context which we
often have to glean from physical inspection.

I'm objecting to the OpenBSD code because it gives the impression of
order and structure, where none exists.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.