* respond to winit platform theme changed event and send it to the layout thread
Signed-off-by: Lloyd Massiah <artmis9@protonmail.com>
* refactoring viewport and theme change handling functions based on feedback
Signed-off-by: Lloyd Massiah <artmis9@protonmail.com>
* fixing issues reported by test-tidy
Signed-off-by: Lloyd Massiah <artmis9@protonmail.com>
* update stylo in order to use color_scheme function on Device
Signed-off-by: Lloyd Massiah <artmis9@protonmail.com>
---------
Signed-off-by: Lloyd Massiah <artmis9@protonmail.com>
Co-authored-by: lazypassion <25536767+lazypassion@users.noreply.github.com>
Servo was previously using the inverse of the correct scale factor
which could cause the scrolling speed to be 4x too slow on a machine
with scale factor of 2.0
Signed-off-by: Nico Burns <nico@nicoburns.com>
There is a command-line argument to override the default window size,
but not one for overriding the default screen resolution. This is
important for testing pages that use screen size to have different
behavior.
In addition to adding the new option this change:
- Renames the `--resolution` command-line argument to `--window-size`
to remove ambiguity with the `--screen-size` argument.
- Passes the screen size as device independent (device pixels scaled by
HiDPI factor) to Servo internals. Not only it make it simpler to pass
the `--window-size` override, it makes more sense. Different screens
can have different HiDPI factors and these can be different from the
scale of the window. This makes the screen HiDPI factor totally
independent of the one that Servo uses for the window.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Instead of exposing many different kinds of messages to the compositor
that are routed through the constellation, expose a single message type
which can be sent across IPC channels. In addition, this IPC channel and
the route to the crossbeam channel with the compositor is created along
with the `CompositorProxy`, simplifying what needs to be passed around
during pipeline initialization.
Previously, some image updates (from video) were sent over IPC with a
special serialization routine and some were sent via crossbeam channels
(canvas). Now all updates go over the IPC channel `IpcSharedMemory` is
used to avoid serialization penalties. This should improve performance
and reduce copies for video, but add a memory copy overhead for canvas.
This will improve in the future when canvas renders directly into a
texture.
All-in-all this is a simplification which opens the path toward having a
standard compositor API and reduces the number of duplicate messages and
proxying that had to happen in libservo.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Instead of exposing many different kinds of messages to the compositor
that are routed through the constellation, expose a single message type
which can be sent across IPC channels. In addition, this IPC channel and
the route to the crossbeam channel with the compositor is created along
with the `CompositorProxy`, simplifying what needs to be passed around
during pipeline initialization.
Previously, some image updates (from video) were sent over IPC with a
special serialization routine and some were sent via crossbeam channels
(canvas). Now all updates go over the IPC channel `IpcSharedMemory` is
used to avoid serialization penalties. This should improve performance
and reduce copies for video, but add a memory copy overhead for canvas.
This will improve in the future when canvas renders directly into a
texture.
All-in-all this is a simplification which opens the path toward having a
standard compositor API and reduces the number of duplicate messages and
proxying that had to happen in libservo.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
The current implementation has 3 main issues related to HiDPI:
1. When the window moves from a screen with scale factor of 1.5 to one
with 1 and back to 1.5, the minibrowser toolbar actually ends up
being scaled by a factor of 2.25 instead of 1.5. This is because we
currently use the [set_pixels_per_point] method on egui's Context,
but calling this with a value of `ppp` will modify egui's internal
'zoom factor' to be:
```
zoom_factor = ppp / native_points_per_pixel.
```
where `native_points_per_pixel` is the window system scale factor.
The idea is egui can calculate the final scale factor for translating
its logical points to physical pixels as:
```
points_per_pixel = zoom_factor * native_points_per_pixel
```
where zoom_factor is a factor used for Ctrl+Plus, Ctrl+Minus
behaviour. The problem is when we handle the ScaleFactorChanged winit
event due to window moving between screens, the
`native_points_per_pixel` still has the value of the previous
screen's native scaling factor and not the current screen's factor.
This seems to be the case even if we pass the ScaleFactorChanged
event to egui before we call `set_pixels_per_point`.
2. The egui logic for handing Ctrl+Plus, Ctrl+Minus and Ctrl+0 doesn't
interact well with servoshell's device-pixel-ratio CLI argument which
allows the user to override the HiDPI factor. For example, Ctrl+0
will cause egui to reset the zoom_factor to 1.0 instead of the
override we wanted. Another issue is egui's Ctrl+Plus/Ctrl+Minus
will scale the minibrowser in increments of 0.10 whereas
Servo's own page zoom doesn't (it keeps multiplying by 1.1, so the
actual increments are 0.1, 0.21. 0.33 etc)
3. The inital window size calculation on Linux currently assumes a scale
factor of 1.0. This means the window doesn't have the expected
default logical size of 1024*740 on HiDPI systems. On a screen with
HiDPI factor of 1.5, the logical window size ends up being 682x493.
This change addresses all 3 issues:
For 1, switch to the `set_zoom_factor` method of egui context
to avoid the issue with scaling by incorrect native_points_per_pixel.
To allow for the device-pixel-ratio override to work, we calculate the
actual zoom_factor as `device-pixel-ratio / window's scaling factor`.
For 2, disable egui's handling of Ctrl+Plus, Ctrl+Minus, Ctrl-0
shortcuts. It is unclear whether the current behaviour of scaling
both the toolbar and the web page was intentional, or just an accident.
This behaviour is also different from other browser where page zoom
doesn't scale the GUI, so it doesn't seem like a regression to me.
For 3, use LogicalSize type of winit which lets the physical size
calulation to be handled by winit using the windows's actual HiDPI
factor instead of hardcoded 1.0.
[set_pixels_per_point]: 1603f05818/crates/egui/src/context.rs (L1886)
Signed-off-by: Mukilan Thiyagarajan <mukilan@igalia.com>
This is part of the switch from `winapi` to `windows-sys`. `windows-sys` is
maintained by Microsoft, so is more "official." More and more crates are
switching to it.
* servoshell: Move desktop files
Move files related to winit into a desktop module.
This is a preparation to merge the android and ohos apps into
servoshell.
* servoshell: Format imports
* servoshell: Move panic hook into separate file
* servoshell: Move desktop main
* Consider ohos as not desktop
* servoshell: Adjust dependencies for shared code
* servoshell: Remove native-bluetooth from default features
There currently is no good way to have target specific default features.
* Rename desktop_main.rs to cli.rs
* Remove todo
