The previous setting of 10 wasn't 10 mm, it was used against the deltas
normalized to a 1000DPI mouse, i.e. closer to 4mm. It was also also per-event,
so a slow movement or a high-frequency touchpad can struggle to meet the
threshold.
Change the trigger to be ~5 mm from the initial touch down, acc
Signed-off-by: Peter Hutterer
---
src/evdev-mt-touchpad.h | 39 ---
1 file changed, 20 insertions(+), 19 deletions(-)
diff --git a/src/evdev-mt-touchpad.h b/src/evdev-mt-touchpad.h
index 65581ae..8ea7d84 100644
--- a/src/evdev-mt-touchpad.h
+++ b/src/evdev-mt-
Previous code used a device coordinate threshold of 300 which won't work on
Elantech touchpads (1280 vs the ~4000 that synaptics has).
Convert to normalized DPI and reduce the threshold to 3mm.
https://bugs.freedesktop.org/show_bug.cgi?id=89206
[note that because of the MAGIC in the accel code th
Ideally we want to specify various thresholds in mm, but not all touchpads
set the hardware resolutions. Rather than conditions to check for resolutions
everywhere, use a macro to give us a normalized value that we use for motion
as well.
Signed-off-by: Peter Hutterer
---
src/evdev.h | 6 ++
Previous code used a device coordinate threshold of 300 which won't work on
Elantech touchpads (1280 vs the ~4000 that synaptics has).
Convert to normalized DPI and reduce the threshold to 3mm.
https://bugs.freedesktop.org/show_bug.cgi?id=89206
[note that because of the MAGIC in the accel code th
On a semi-mt device lifting slot 0 before slot 1 makes slots 1 become slot 0
(with the matching coordinate jump), potentially triggering the tap movement
threshold.
On one test we can just swap the release order, the other test we need to
disable (the _inverted version of this test tests the other
All callers except the tap motion threshold call
tp_get_delta() followed by tp_filter_motion() - the latter normalized it
before calling into the accleration code.
Move the normalization into tp_get_delta() so we don't deal with
device-specific coordinates but normalized deltas instead.
Signed-of
On 03/05/2015 03:15 AM, Marek Chalupa wrote:
I wonder if this is right. Until now, when you had display_name,
you knew that the socket was $XDG_RUNTIME_DIR/display_name.
Now it can be anywhere and you still have just name of it. However, this
does not look like a big deal since
s->display_name i
On Wed, Mar 4, 2015 at 11:25 PM, Peter Hutterer
wrote:
> Signed-off-by: Peter Hutterer
> ---
For the series:
Reviewed-by: Benjamin Tissoires
> src/evdev-tablet.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/src/evdev-tablet.c b/src/evdev-tablet.c
> index 8c11409..3e3924c 100644
>
On 05/03/15 02:15 AM, Pekka Paalanen wrote:
> On Wed, 4 Mar 2015 16:26:25 -0600
> Derek Foreman wrote:
>
>> Certain circumstances may lead to the "force" clause in
>> input_set_pointer_image() being reached when the current cursor
>> is blank or unset. These are special cursors that don't have
Le 03/03/2015 13:49, Davide Bettio a écrit :
> Signed-off-by: Davide Bettio
> ---
> doc/man/wl_display_connect.xml | 8
> src/wayland-client.c | 22 +-
> src/wayland-server.c | 23 ++-
> 3 files changed, 39 insertions(+), 14 de
On Tue, Mar 3, 2015 at 7:49 AM, Davide Bettio
wrote:
> Signed-off-by: Davide Bettio
> ---
> src/wayland-server.c | 17 ++---
> 1 file changed, 10 insertions(+), 7 deletions(-)
>
> diff --git a/src/wayland-server.c b/src/wayland-server.c
> index 05e9c07..2922e41 100644
> --- a/src/wa
On Tue, Mar 3, 2015 at 7:49 AM, Davide Bettio
wrote:
> Signed-off-by: Davide Bettio
> ---
> doc/man/wl_display_connect.xml | 8
> src/wayland-client.c | 22 +-
> src/wayland-server.c | 23 ++-
> 3 files changed, 39 insertions
On Mon, Mar 02, 2015 at 10:38:54AM -0500, Benjamin Tissoires wrote:
> On Mon, Mar 2, 2015 at 1:53 AM, Peter Hutterer
> wrote:
> > On Thu, Feb 26, 2015 at 04:20:49PM -0500, Benjamin Tissoires wrote:
> >> On Tue, Feb 24, 2015 at 1:21 AM, Peter Hutterer
> >> wrote:
> >> > Needs to be calculated fro
Hi,
On 05-03-15 04:29, Peter Hutterer wrote:
On Wed, Mar 04, 2015 at 03:26:30PM +0100, Hans de Goede wrote:
Extend the touchpad gesture API with zoom/rotate gestures. Note that this
new API offers a single event stream for both zoom and rotate data, this
is deliberate as some applications may b
On Tue, 27 Jan 2015 12:13:32 -0800
Bryce Harrington wrote:
> On Mon, Oct 13, 2014 at 07:40:51PM +0200, Niels Ole Salscheider wrote:
> > You can now use "C" to switch between the following modes:
> > - Assume sRGB input color space
> > - Assume that the input color space is the blending color spac
On Wed, 4 Mar 2015 16:26:25 -0600
Derek Foreman wrote:
> Certain circumstances may lead to the "force" clause in
> input_set_pointer_image() being reached when the current cursor
> is blank or unset. These are special cursors that don't have
> images, and they need to be handled differently tha
17 matches
Mail list logo
