On Tue, Mar 12, 2013 at 7:07 PM, Christophe Fergeau <cfergeau@xxxxxxxxxx> wrote: > On Tue, Mar 12, 2013 at 06:28:36PM +0200, Zeeshan Ali (Khattak) wrote: >> On Tue, Mar 12, 2013 at 3:16 PM, Christophe Fergeau <cfergeau@xxxxxxxxxx> wrote: >> > On Sat, Mar 09, 2013 at 04:44:14PM +0200, Zeeshan Ali (Khattak) wrote: >> >> From: "Zeeshan Ali (Khattak)" <zeeshanak@xxxxxxxxx> >> >> >> >> This patch adds API to set/get image compression options on >> >> domain/graphics[@type='spice'] nodes. >> >> >> >> Also included are simple tests for this API. >> > >> > Do we really need a dedicated gobject for this? I would have gone with >> > gvir_config_domain_graphics_spice_[gs]et_image_compression() >> >> It would have been easier to go w/o a separate object but I went this >> way since 'image' is a separate node in the config and I would assume >> its for a reason. One reason that comes to mind is that in future >> there might be some other options being added here? > > We don't always go with 'one node' = 'one class', see the <os> node for > example, or gvir_config_domain_disk_set_startup_policy (various other > similar examples in this case). > For xml nodes that are very simple, they generally are modelled along with > the parent node, we can always deprecate the simple setter in the parent > node if the node gets more complex in the future. > It's something to decide on a case by case basis, but here I'd tend to go > with the simpler approach. Ah ok, makes sense. I was under the impression that we always followed 'one node' = 'one class' rule. I'll change this part as well then. Thanks. -- Regards, Zeeshan Ali (Khattak) FSF member#5124 -- libvir-list mailing list libvir-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libvir-list