Re: [PATCH wayland v2] protocol: Clarify the behaviour of wl_surface.attach

2017-02-16 Thread Pekka Paalanen
On Wed, 15 Feb 2017 11:49:33 -0600 Derek Foreman wrote: > On 15/02/17 06:50 AM, Pekka Paalanen wrote: > > On Wed, 15 Feb 2017 12:17:10 +0800 > > Jonas Ådahl wrote: > > > >> On Tue, Feb 14, 2017 at 10:20:06AM -0600, Derek Foreman wrote: > >>> Attaching a NULL wl_buffer to a surface is not alw

Re: [PATCH wayland v2] protocol: Clarify the behaviour of wl_surface.attach

2017-02-15 Thread Derek Foreman
On 15/02/17 06:50 AM, Pekka Paalanen wrote: On Wed, 15 Feb 2017 12:17:10 +0800 Jonas Ådahl wrote: On Tue, Feb 14, 2017 at 10:20:06AM -0600, Derek Foreman wrote: Attaching a NULL wl_buffer to a surface is not always valid, but the previous text indicated it was. Instead, let's define what NUL

Re: [PATCH wayland v2] protocol: Clarify the behaviour of wl_surface.attach

2017-02-15 Thread Derek Foreman
On 14/02/17 10:17 PM, Jonas Ådahl wrote: On Tue, Feb 14, 2017 at 10:20:06AM -0600, Derek Foreman wrote: Attaching a NULL wl_buffer to a surface is not always valid, but the previous text indicated it was. Instead, let's define what NULL attachment does for all the surface roles defined in wayla

Re: [PATCH wayland v2] protocol: Clarify the behaviour of wl_surface.attach

2017-02-15 Thread Mike Blumenkrantz
On Tue, Feb 14, 2017 at 11:17 PM Jonas Ådahl wrote: > On Tue, Feb 14, 2017 at 10:20:06AM -0600, Derek Foreman wrote: > > Attaching a NULL wl_buffer to a surface is not always valid, but > > the previous text indicated it was. > > > > Instead, let's define what NULL attachment does for all the sur

Re: [PATCH wayland v2] protocol: Clarify the behaviour of wl_surface.attach

2017-02-15 Thread Pekka Paalanen
On Wed, 15 Feb 2017 12:17:10 +0800 Jonas Ådahl wrote: > On Tue, Feb 14, 2017 at 10:20:06AM -0600, Derek Foreman wrote: > > Attaching a NULL wl_buffer to a surface is not always valid, but > > the previous text indicated it was. > > > > Instead, let's define what NULL attachment does for all the

Re: [PATCH wayland v2] protocol: Clarify the behaviour of wl_surface.attach

2017-02-14 Thread Jonas Ådahl
On Tue, Feb 14, 2017 at 10:20:06AM -0600, Derek Foreman wrote: > Attaching a NULL wl_buffer to a surface is not always valid, but > the previous text indicated it was. > > Instead, let's define what NULL attachment does for all the surface > roles defined in wayland.xml, stop giving a blanket defi

[PATCH wayland v2] protocol: Clarify the behaviour of wl_surface.attach

2017-02-14 Thread Derek Foreman
Attaching a NULL wl_buffer to a surface is not always valid, but the previous text indicated it was. Instead, let's define what NULL attachment does for all the surface roles defined in wayland.xml, stop giving a blanket definition of its behavior in wl_surface.attach, and warn developers that the