On Wed, Mar 19, 2014 at 01:54:13AM +0000, Wang, Quanxian wrote: > [Wang, Quanxian] Thanks Bryce. > >I glanced at the other patches and they look technically correct, but I > >think what > >might be missing here is a stronger justification about why this is needed?
> [Wang, Quanxian] name is one property of output which is assigned > detected by backend. In compositor, it is visible. Why is it assigned > as a name in compositor? I think display it in compositor logs will be > one of reason. Therefore the same thing in client. At the same time, > it provides a good communication with compositor as a choice. Client > could take it as a parameter to communicate with compositor, instead > of only by wl_output object. I like your word 'human-readable' > especially for client. Give client a choice to communicate with > compositor without wl_output object, and also give client a choice to > display human-readable name in log which is visible to > customer. Currently I am designing weston randr protocol, > customer(admin or developer) could input a char name to stand for an > output from command line. It is just one usage. Anyway, name of output That's a good example to include for justification; maybe include mention of the randr use case in your commit message. > should be visible for both of compositor and client. > > >> + </description> > >> + <arg name="name" type="string" summary="name of output"/> > >> + </event> > >> </interface> > >> > >> <interface name="wl_region" version="1"> > >> -- > >> 1.8.1.2 > >> > >> _______________________________________________ > >> wayland-devel mailing list > >> wayland-devel@lists.freedesktop.org > >> http://lists.freedesktop.org/mailman/listinfo/wayland-devel _______________________________________________ wayland-devel mailing list wayland-devel@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/wayland-devel