[Nut-upsdev] Re: [nut-commits] svn commit r831 - in trunk: . docs scripts

[Arjen de Korte]

Peter Selinger wrote:

> However, if the user specifies a list of drivers with
> --with-drivers=driver[,driver...], then it is not so easy to detect
> whether any of these require USB support. For this, the configure
> script would have to contain a list of USB drivers, which adds one
> more thing to update that could easily be forgotten.

Hmmm, that's another way to look at it.

> Is it a problem to install the hotplug/udev rules, even when no USB
> driver has been requested?

Partially. The rules assume that the user specified in RUN_AS_USER
exists, which may not be (and in my case is not) the case. Therefor at
startup the system will complain about about this.

> By default, this will only be done if the corresponding directory
> exists, so should not give any errors or strange behavior.

Other than the above, no problems.

> The rule will never do anything unless a known UPS
> is plugged into the USB port, in which case it's probably desired. 

I agree. Yet it is a little counter intuitive that if you don't want USB
support compiled in (for whatever reason) and specify --without-usb, you
still have to add --without-hotplug-dir and --without-udev-dir to get
rid of the startup errors.

I had expected that --without-usb implies --without-hotplug-dir and
--without-udev-dir.

Best regards, Arjen