Hi Yan,
thanks for the detailed reply, that was very helpful. It is also good to
see that someone else has tried this before.
I'll definitely give this a go.
Iago
On lun, 2013-07-08 at 23:13 -0700, yan.w...@linux.intel.com wrote:
> Hi,
> I have implemented Wayland buffer sharing mechanism in
Hi,
I have implemented Wayland buffer sharing mechanism in WebKit2-efl based
on nested client example. Nested client share buffer from one nested
client to nesting client which is the Wayland server of nested client
and Wayland client of Weston at the same time.
For WebKit2, there may be no onl
On 7/9/2013 3:47 AM, Quentin Glidic wrote:
> From: Quentin Glidic
>
> Long options with argument support two formats: equal ("--long=arg") and
> space ("--long arg")
> Short options now support three formats: short ("-sarg"), equal
> ("-s=arg") and space ("-s value")
>
> Provide a test program
>
There are two character widths.
The terminal should be using the string width reported from cairo for
all graphics, such as the underscore shown here. A correct
implementation should work acceptably with proportionally-spaced fonts.
This is certainly not true of the current wayland terminal.
Though I agree that the built-in default should load the wlshm driver if
needed, this is not the problem I am trying to fix.
My /etc/X11/xorg.conf file contains 'Driver "nv"'. This causes the
wayland xserver to fail. This will happen whether or not wlshm or wlegl
drivers are needed. I suspect
A similar function is wcwidth(). There is an implementation here:
http://www.cl.cam.ac.uk/~mgk25/ucs/wcwidth.c
I don't know how it compares to the complexity or feature completeness
of Pango's implementation (which I think is a wrapper around HarfBuzz)
On Mon, Jul 8, 2013 at 5:07 PM, Kristian Hø
From: Quentin Glidic
Signed-off-by: Quentin Glidic
---
man/weston.man | 6 +++---
shared/config-parser.h | 1 +
shared/option-parser.c | 15 +++
src/compositor.c | 31 ++-
tests/weston-tests-env | 6 --
5 files changed, 49 insertions(
From: Quentin Glidic
Long options with argument support two formats: equal ("--long=arg") and
space ("--long arg")
Short options now support three formats: short ("-sarg"), equal
("-s=arg") and space ("-s value")
Provide a test program
Signed-off-by: Quentin Glidic
---
man/weston.man
I've commented on this before: the fix here isn't to introduce a new
config file to worry about, the fix is to make the X server have a
different fallback driver for when it's running in xwayland mode.
Kristian
On Wed, Jul 3, 2013 at 4:19 AM, Bill Spitzak wrote:
> I am really interested in knowi
On Tue, Jul 09, 2013 at 12:06:24AM +0200, Quentin Glidic wrote:
> From: Quentin Glidic
That's better, thanks.
Kristian
> Signed-off-by: Quentin Glidic
> ---
> shared/Makefile.am | 11 ---
> tests/.gitignore | 1 +
> tests/Makefile.am
On Mon, Jul 08, 2013 at 06:59:45PM +0100, Rob Bradford wrote:
> From: Rob Bradford
>
> The RDP compositor dynamically creates and destroys seats for the
> different connecting clients. Althrough the memory for the seat was
> being freed and it was also being removed from the list of seats in the
On Thu, Jun 06, 2013 at 03:32:42PM +0800, Peng Wu wrote:
> ---
> clients/Makefile.am | 2 +-
> clients/terminal.c | 18 --
> 2 files changed, 17 insertions(+), 3 deletions(-)
This makes sense, but I'm not happy about the pango dependency. One
of the few reasons (if not the only
From: Quentin Glidic
Signed-off-by: Quentin Glidic
---
shared/Makefile.am | 11 ---
tests/.gitignore | 1 +
tests/Makefile.am | 12 +++-
{shared => tests}/config-parser-test.c | 0
4 files changed, 12 insertions(+)
This average will result in a value that does not match any of the
characters. Using the width of '2' probably works better. This means
that at least digits will line up (it is very common for all digits
except perhaps '1' to have the same width). In theory you can use 'n'
but I have gotten bet
On Thu, Jun 06, 2013 at 03:32:41PM +0800, Peng Wu wrote:
> ---
> clients/terminal.c | 55
> --
> 1 file changed, 41 insertions(+), 14 deletions(-)
Thanks, that makes sense, committed. Sorry for the delay.
Kristian
> diff --git a/clients/term
On Mon, Jul 8, 2013 at 11:56 AM, Rob Bradford wrote:
> If we're talking changes...
>
> Let's make the popup return an object with a done event on rather than
> stick the event in the interface like popup_done is.
So, you want that set_popup automatically sending a done event after
called? Then se
Jason Ekstrand wrote:
The other option here is to specify that the client does not get the
leave event if the surface is destroyed.
I greatly prefer this solution to this sort of problem.
___
wayland-devel mailing list
wayland-devel@lists.freedeskto
From: Rob Bradford
The RDP compositor dynamically creates and destroys seats for the
different connecting clients. Althrough the memory for the seat was
being freed and it was also being removed from the list of seats in the
compositor the client was not being informed of that the seat was no
lon
From: Rafael Antognolli
xdg_shell is a protocol aimed to substitute wl_shell in the long term,
but will not be part of the wayland core protocol. It starts as a
non-stable API, aimed to be used as a development place at first, and
once features are defined as required by several desktop shells, w
On Mon, Jul 08, 2013 at 01:51:34PM -0400, Kristian Høgsberg wrote:
> On Mon, Jul 08, 2013 at 12:10:54PM +0100, Rob Bradford wrote:
> > *bump*
> >
> > This is a fix for:
> > https://bugs.freedesktop.org/show_bug.cgi?id=65726&list_id=320167
> >
> > Although it might be much better if we could emit
On Mon, Jul 08, 2013 at 12:10:54PM +0100, Rob Bradford wrote:
> *bump*
>
> This is a fix for:
> https://bugs.freedesktop.org/show_bug.cgi?id=65726&list_id=320167
>
> Although it might be much better if we could emit the event
> (especially the leave) before we destroy the surface - that being sai
On Mon, Jul 08, 2013 at 04:15:07PM +0200, Giulio Camuffo wrote:
> There is not much else to say. surfaces created with weston_surface_create
> have
> a NULL resource, and if that gets picked by default_grab_focus (in input.c)
> the next
> seat_get_pointer will break.
>
> I guess another solution c
On Mon, Jul 08, 2013 at 03:32:58PM +0200, Quentin Glidic wrote:
> From: Quentin Glidic
>
> Signed-off-by: Quentin Glidic
Thanks, nice work Quentin, committed.
Kristian
> ---
> configure.ac| 2 +-
> src/weston-launch.c | 4
> 2 files changed, 5 insertions(+), 1 deletion(-)
>
> d
On Mon, Jul 8, 2013 at 11:32 AM, Rob Bradford wrote:
> On 8 July 2013 17:25, Jason Ekstrand wrote:
>
> > That's exactly what weston *was* doing. However, thanks to the destroyed
> > proxies, the clients were getting NULL anyway. We could to go through a
> > bunch of trouble to keep the resource
On 8 July 2013 17:25, Jason Ekstrand wrote:
> That's exactly what weston *was* doing. However, thanks to the destroyed
> proxies, the clients were getting NULL anyway. We could to go through a
> bunch of trouble to keep the resource valid and call wl_pointer.leave on a
> valid resource. Howeve
On Mon, Jul 8, 2013 at 6:10 AM, Rob Bradford wrote:
> *bump*
>
> This is a fix for:
> https://bugs.freedesktop.org/show_bug.cgi?id=65726&list_id=320167
>
> Although it might be much better if we could emit the event
> (especially the leave) before we destroy the surface - that being said
> the des
On Mon, Jul 8, 2013 at 9:52 AM, Rob Bradford wrote:
> On 8 July 2013 15:15, Giulio Camuffo wrote:
>
> Hi Giulio,
>
> > There is not much else to say. surfaces created with
> weston_surface_create
> > have
> > a NULL resource, and if that gets picked by default_grab_focus (in
> input.c)
> > the ne
If we're talking changes...
Let's make the popup return an object with a done event on rather than
stick the event in the interface like popup_done is.
I'd also like to see us standardise on the naming of the request to
create the new object - I think create_xdg_surface is better than
get_xdg_sur
On 8 July 2013 15:15, Giulio Camuffo wrote:
Hi Giulio,
> There is not much else to say. surfaces created with weston_surface_create
> have
> a NULL resource, and if that gets picked by default_grab_focus (in input.c)
> the next
> seat_get_pointer will break.
I guess my point is that I wanted yo
On 04/07/2013 22:20, Kristian Høgsberg wrote:
On Thu, Jul 4, 2013 at 12:35 PM,
wrote:
And they must link directly! That is the proper way to do that.
Providing API using these libraries does not mean forcing anyone to
use that. And using a library does require linking to it. Hiding
this depend
On 05/07/2013 20:10, Bill Spitzak wrote:
On 07/03/2013 03:35 PM, scsijon wrote:
we have seen this before with xorg servers, it's usually found that some
dependancy for the driver concerned has been updated, but the driver
itself hasn't been rebuilt to suit, usually fixed by either recompiling
t
On Mon, Jul 8, 2013 at 3:40 PM, Iago Toral wrote:
> El 2013-07-08 08:38, Jonas Ådahl escribió:
>
>> On Mon, Jul 8, 2013 at 8:05 AM, Iago Toral wrote:
>>>
>>> Hi,
>>>
>>> I am working on porting WebKitGTK+ to Wayland and we are having some
>>> difficulties figuring out the proper way to deal with
There is not much else to say. surfaces created with weston_surface_create
have
a NULL resource, and if that gets picked by default_grab_focus (in input.c)
the next
seat_get_pointer will break.
I guess another solution could be to make weston_compositor_pick_surface not
pick surfaces without a val
OK, will send an updated patch later.
On Fri, Jul 5, 2013 at 6:09 PM, Kristian Høgsberg wrote:
> Just a quick comment - I've been itching to rename wl_shell_surface so
> something more concise, so I'd like to go with xdg_surface instead of
> xdg_shell_surface.
>
> Kristian
>
> On Fri, Jul 5, 2013
Can you provide some more explanation in your commit message about why
the pointer might have a surface focussed but that surface does not
have a valid resource. (I'm wondering if this is fixing the symptom of
a problem elsewhere.)
Rob
On 7 July 2013 16:38, Giulio Camuffo wrote:
> seat->pointer-
Looks good to me.
--Jason Ekstrand
On Jul 7, 2013 10:41 AM, "Giulio Camuffo" wrote:
> seat->pointer->focus->resource can be NULL, if the surface was
> created with weston_surface_create.
> ---
> src/input.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/src/input.c b/sr
From: Quentin Glidic
Signed-off-by: Quentin Glidic
---
configure.ac| 2 +-
src/weston-launch.c | 4
2 files changed, 5 insertions(+), 1 deletion(-)
diff --git a/configure.ac b/configure.ac
index 726eb4f..0864d10 100644
--- a/configure.ac
+++ b/configure.ac
@@ -52,7 +52,7 @@ AC_CHE
El 2013-07-08 08:38, Jonas Ådahl escribió:
On Mon, Jul 8, 2013 at 8:05 AM, Iago Toral wrote:
Hi,
I am working on porting WebKitGTK+ to Wayland and we are having some
difficulties figuring out the proper way to deal with the
multiprocess
architecture introduced with WebKit2.
In WebKit2 we ha
From: Quentin Glidic
We can then add tests which do not use Weston in the test suite.
Signed-off-by: Quentin Glidic
---
tests/.gitignore | 7 +--
tests/Makefile.am | 41 ++---
2 files changed, 23 insertions(+), 25 deletions(-)
diff --git a/tests/.giti
*bump*
This is a fix for:
https://bugs.freedesktop.org/show_bug.cgi?id=65726&list_id=320167
Although it might be much better if we could emit the event
(especially the leave) before we destroy the surface - that being said
the destruction of the object will also result in the proxy being
removed
Hi Pekka,
You raise a good point. I don't think the increase in complexity is
worth it - we have good way to construct a surface list from the
layers (and we do it at the start of every frame) anyway: I think it's
nice that the management of this cache is really encapsulated in a
single place.
Ro
On Mon, Jul 8, 2013 at 8:05 AM, Iago Toral wrote:
> Hi,
>
> I am working on porting WebKitGTK+ to Wayland and we are having some
> difficulties figuring out the proper way to deal with the multiprocess
> architecture introduced with WebKit2.
>
> In WebKit2 we have two processes that are responsibl
42 matches
Mail list logo
