On 09/30/2009 02:15 PM, David Lutterkort wrote:
On Tue, 2009-09-29 at 14:39 -0400, Laine Stump wrote:
> The originalpatch for ncf_if_xml_state would search for the first ip
> node within the ipv4 protocol node, and replace it with the results of
> SIOCGIFADDR (or sometimes just add a new one on.)
>
> Now we remove all ip addresses sourced from the config, even if we
> can't get any replacements from the device.
>
The big question (as I mentioned in the other mail) is: what does
ncf_if_xml_state report ? Only live information or a combination of live
+ configured info ? I could be persuaded either way - it mostly depends
on how people will use the results of ncf_if_xml_state: is ist mostly
for information/reporting, or should it be convenient to use the result
of ncf_if_xml_state and pass it to ncf_define ?
I must confess that I went with the former (live info only) at least
partly because it was easiest to make that work with libvirt in the
short term (libvirt parses the xml, and currently throws away all but
the first ip node; that will need to be changed anyway, of course).
From a complete vaccum of use cases, I'm assuming that ncf_if_xml_state
will be used mostly for information/reporting; I may be completely
wrong, though. At any rate, I still don't have a strong enough opinion
to try and convince you to cross over though. Hopefully someone with a
strong opinion (and a good reason for it) can step in.