Fix 998529, mono (invert) cursors not visible on a black background, by doing
simple edge detection on the cursor (this is done once when the cursor
is changed and then cached, cursors are 32x32 generally) and thus having
a cursor with contrast on both dark and light backgrounds.
When (if) GDK
On 10/18/2013 02:30 PM, Alon Levy wrote:
Please ignore, I'll send a new version.
Fix 998529, mono (invert) cursors not visible on a black background, by doing
simple edge detection on the cursor (this is done once when the cursor
is changed and then cached, cursors are 32x32 generally) and thus
---
gtk/channel-cursor.c | 32
1 file changed, 32 insertions(+)
diff --git a/gtk/channel-cursor.c b/gtk/channel-cursor.c
index d1d2c34..66e91df 100644
--- a/gtk/channel-cursor.c
+++ b/gtk/channel-cursor.c
@@ -246,6 +246,35 @@ static void
Fix 998529, mono (invert) cursors not visible on a black background, by doing
simple edge detection on the cursor (this is done once when the cursor
is changed and then cached, cursors are 32x32 generally) and thus having
a cursor with contrast on both dark and light backgrounds.
When (if) GDK
Some versions of gcc warn about:
red_channel.c: In function 'red_channel_client_wait_outgoing_item':
red_channel.c:2331: error: 'end_time' may be used uninitialized in this
function [-Wuninitialized]
red_channel.c: In function 'red_channel_client_wait_pipe_item_sent':
red_channel.c:2363: error:
Hey,
ACK and pushed, sorry for the late answer. This was broken with 9a62a9a809
as this code is only conditionally compiled.
Christophe
On Sat, Oct 12, 2013 at 05:56:00PM +0800, Axel Lin wrote:
Fix missing monitor_latency argument in red_channel_client_create call.
Signed-off-by: Axel Lin
On Sat, Oct 12, 2013 at 05:56:00PM +0800, Axel Lin wrote:
Fix missing monitor_latency argument in red_channel_client_create call.
Fwiw, it would probably be better if we just killed red_tunnel_worker.c as
I don't think anyone is using or testing it. Do you have a use for it, or
do you just
On 10/14/2013 05:52 PM, Fabio Fantoni wrote:
/ I have added vdagent support on libxl for hvm domUs using upstream qemu.
// On windows domUs it is working without problems found in over one year
// that I'm using it.
//
// On linux domUs is not working, it seems to be a problem on the kernel
2013/10/18 Christophe Fergeau cferg...@redhat.com:
On Sat, Oct 12, 2013 at 05:56:00PM +0800, Axel Lin wrote:
Fix missing monitor_latency argument in red_channel_client_create call.
Fwiw, it would probably be better if we just killed red_tunnel_worker.c as
I don't think anyone is using or
- Original Message -
What are guest OS, agent driver, and client versions?
guest is fedora19, with last updates
client :
remote-viewer : 0.5.7
spice-gtk3 0.21 (same with 0.20)
Can you provide SPICE_DEBUG=1 log of spice client?
(G_MESSAGES_DEBUG=all SPICE_DEBUG=1
- Original Message -
hi List,
I get a no map for 173 error when touching the minus key on recent browsers
and on a spanish browser.
I know there are known issues regarding translation of web key codes into
scan codes and am eager to help solving this , a precedent thread pointed
well that s weird, im using firefox too
and add the mapping there, but wouldnt work. the following change
solves the problem for me though
firefox_scanmap[173] = 0x4a;
regards,
On 10/18/2013 05:00 PM, Marc-André
Lureau wrote:
- Original Message -
well that s weird, im using firefox too
and add the mapping there, but wouldnt work. the following change solves
the problem for me though
/firefox_scanmap[173] = 0x4a;/
Ah, then it was fixed some time ago:
- Original Message -
2013/10/18 Christophe Fergeau cferg...@redhat.com:
On Sat, Oct 12, 2013 at 05:56:00PM +0800, Axel Lin wrote:
Fix missing monitor_latency argument in red_channel_client_create call.
Fwiw, it would probably be better if we just killed red_tunnel_worker.c as
Hey,
On Wed, Oct 16, 2013 at 08:45:31AM -0500, Jeremy White wrote:
On 10/16/2013 06:05 AM, Christophe Fergeau wrote:
This does not seem right, snd_codec_create() hardcodes the numbers of
channels to SND_CODEC_CELT_PLAYBACK_CHAN, so we totally ignore whatever was
in 'start', but here we use
I'm willing to be persuaded otherwise.
As said on IRC, I'm not looking for something fully dynamic, but if
more of the number of channel/sample size could be hidden in
SndCodec (while still being hardcoded), this would look nicer imo.
But I agree it only makes sense if it's straightforward
- Original Message -
well that s weird, im using firefox too and add the mapping there, but
wouldnt work. the following change solves the problem for me though
/firefox_scanmap[173] = 0x4a;/
Ah, then it was fixed some time ago:
Besides, I still not understand how that should work. You (Marc) told
me that it is easy to get the correct scancodes, but the sources
contain the following comment:
Did I? I don't remember the context.
Again, my keysym patches provides a workaround for that problem.
** This will
- Original Message -
Besides, I still not understand how that should work. You (Marc) told
me that it is easy to get the correct scancodes, but the sources
contain the following comment:
Did I? I don't remember the context.
Again, my keysym patches provides a workaround
I understand. you are sending keysym, and avoid the layout and hw scancode...
But Spice, in general, is talking to a VM, and currently only via scancodes.
Quite clearly, I understand that using the agent and XTest, input-methdod or
accessible-technology will avoid the hw layer altogether. We
20 matches
Mail list logo
