On Tue, Jan 03, 2012 at 10:08:10PM +0000, Daniel P. Berrange wrote:
On Tue, Jan 03, 2012 at 11:38:15PM +0200, Zeeshan Ali (Khattak)
wrote:
> On Tue, Jan 3, 2012 at 12:11 PM, Christophe Fergeau <cfergeau(a)redhat.com>
wrote:
> > There are some devices which we are not able to parse yet, in these
> > cases a runtime warning is printed to tell what's going on.
>
> Actually I see that most of the device types are not being handled.
> I would suggest we implement all the cases before merging this or
> otherwise we'll easily forget about this. Yes the warning won't let us
> forget but I don't think emitting a warning for each unimplemented
> type (by the library itself) is a good idea since these warnings are
> visible to users/app developers and they are useless for them.
Agreed, the way we deal with the XML docs ensures we preserve
any parts of the DOM we don't know about. IMHO we should only
emit a warning if there is a bit of the DOM we don't expect
and we are loosing that data during the requested manipulation.
I'll change these warnings to g_debug then
Christophe