It took a long time to check this problem.
Finally, I found that due to a detail missing when changing #34832, the
bounds of `ScrollHandle` was not updated in the Interactivity `prepaint`
phase.
```diff
- scroll_handle_state.padded_content_size = padded_content_size;
+ scroll_handle_state.max_offset = scroll_max;
```
It was correct before the change, because the `padded_content_size`
(including `bounds.size`) was saved before, and the bounds was missing
after changing to `max_offset`, but the bounds were not updated
anywhere.
So when `scroll_handle.bounds()` is obtained outside, it is always 0px
here.
@MrSubidubi
Release Notes:
- N/A
Release Notes:
- N/A
Just make a simple change to avoid crash.
```
thread 'main' panicked at library\std\src\time.rs:436:33:
overflow when subtracting duration from instant
stack backtrace:
0: std::panicking::begin_panic_handler
at /rustc/17067e9ac6d7ecb70e50f92c1944e545188d2359/library\std\src\panicking.rs:697
1: core::panicking::panic_fmt
at /rustc/17067e9ac6d7ecb70e50f92c1944e545188d2359/library\core\src\panicking.rs:75
2: core::panicking::panic_display
at /rustc/17067e9ac6d7ecb70e50f92c1944e545188d2359/library\core\src\panicking.rs:261
3: core::option::expect_failed
at /rustc/17067e9ac6d7ecb70e50f92c1944e545188d2359/library\core\src\option.rs:2024
4: core::option::Option::expect
at /rustc/17067e9ac6d7ecb70e50f92c1944e545188d2359/library\core\src\option.rs:933
5: std::time::impl$3::sub
at /rustc/17067e9ac6d7ecb70e50f92c1944e545188d2359/library\std\src\time.rs:436
6: data_table::Quote::random
at .\crates\gpui\examples\data_table.rs:54
```
Release Notes:
- N/A
---
This change is used to solve the problem of not being able to respond
correctly in two-layer scrolling (in different directions). This is a
common practical requirement.
As in the example, in actual use, there may be a scene with a horizontal
scroll in a vertical scroll. Before the modification, if we scroll up
and down in the area that can scroll horizontally, it will not respond
(because it is blocked by the horizontal scroll layer).
## Before
https://github.com/user-attachments/assets/e8ea0118-52a5-44d8-b419-639d4b6c0793
## After
https://github.com/user-attachments/assets/aa14ddd7-5596-4dc5-9c6e-278aabdfef8e
----
This change may cause many side effects, causing some scrolling details
to be different from before, and more testing and analysis are needed.
I have tested some existing scenarios of Zed (such as opening the Branch
panel on the Editor and scrolling) and it seems to be correct (but it is
possible that I don’t know some interaction details). Here, the person
who added this line of code before needs to evaluate the original
purpose.
Release Notes:
- N/A
----
The before version GPUI used `Cancel` for cancel text, if we use
non-English text (e.g.: "取消" in Chinese), then the press `Esc` to cancel
will not work.
So this PR to change it by use `PromptButton` to instead the `&str`,
then we can use `PromptButton::cancel("取消")` for the `Cancel` button.
Run `cargo run -p gpui --example window` to test.
---
Platform Test:
- [x] macOS
- [x] Windows
- [x] Linux (x11 and Wayland)
---------
Co-authored-by: Mikayla Maki <mikayla@zed.dev>
Co-authored-by: Mikayla Maki <mikayla.c.maki@gmail.com>
Reverts zed-industries/zed#31022
Sorry @mikayla-maki, I found that things are more complicated than I
thought.
The lines returned by shape_text must maintain the same length as all
the original characters, otherwise the subsequent offset needs to always
consider the difference of `\r\n` or `\n` to do the offset.
Before, we only needed to add +1 after each offset after the line, but
now we need to consider +1 or +2, which is much more complicated.
Release Notes:
- N/A
---
Today I check the shape_text result on Windows, I get:
<img width="409" alt="屏幕截图 2025-05-20 222908"
src="https://github.com/user-attachments/assets/3ee93911-3de1-4e01-9433-00c626fc2369"
/>
Here the `shape_text` split logic I think it should use `lines` method,
not `split('\n')`, the newline on Windows is `\r\n`.
Release Notes:
- N/A
---
When we use `window_handle` to draw WebView on Windows, this will crash
by:
This error caused by when used WebView2.
```
thread 'main' panicked at crates\gpui\src\app\async_context.rs:91:28:
already borrowed: BorrowMutError
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
thread 'main' panicked at library\core\src\panicking.rs:221:5:
panic in a function that cannot unwind
```
Try this https://github.com/tauri-apps/wry/pull/1383 on Windows can
replay the crash.
In fact, we had done [a similar fix around August last
year](https://github.com/huacnlee/zed/pull/6), but we used the unsafe
method to avoid crashes in that version, we felt that it was not a good
change, so we do not make PR.
Today @sunli829 thought about it again and changed the method. Now using
`try_borrow_mut` is similar to the previous `borrow_mut`.
https://github.com/zed-industries/zed/blob/691de6b4b36d3ada91a1e238904b065eec454188/crates/gpui/src/app.rs#L70-L78
I have tested to start Zed by those changes, it is looks no problem.
Co-authored-by: Sunli <scott_s829@163.com>
Release Notes:
- N/A
---
Fix this long-standing issue so that we can support Link hover colors.
And renamed `text_layout` example to `text_style`.
---
I spent some time studying the process of this text style change and
found it a bit complicated.
At first, I thought there was a problem with refine and it was not
passed properly. After changing it, I found that it was not the problem.
Then I found that it was because `TextRun` had already stored the
`color`, `background`, `underline`, `strikethrough` in TextRun in the
`request_layout` stage. They area calculate at the `request_layout`
stage, but request_layout stage there was no `hitbox`, so the hover
state was not obtained.
```bash
cargo run -p gpui --example text_style
```
https://github.com/user-attachments/assets/24f88f73-775e-41d3-a502-75a7a39ac82b
---------
Co-authored-by: Mikayla Maki <mikayla.c.maki@gmail.com>
Release Notes:
- Improved performance when using the scroll wheel and some other mouse
interactions.
Based on some cache details about GPUI `AnyView::cached` that I found in
the discussion of
https://github.com/zed-industries/zed/discussions/24260#discussioncomment-12135749,
and combined with the optimization points found in actual applications.
This change may have some scenarios that I have not considered, so I
just make a draft to put forward my ideas first for discussion.
From my analysis, `AnyView::cached` will always invalid by Div's mouse
events, because of it called `window.refresh`. I understand that (mouse
move event) this is because the interface changes related to hover and
mouse_move will be affected style, so `window.refresh` is required.
Since Div does not have the `entity_id` of View, it is impossible to
know which View should be refreshed, so the entire window can only be
refreshed.
With this change, we can reduce a lot of `render` method calls on
ScrollWheel or Mouse Event.
Release Notes:
- N/A
As https://github.com/zed-industries/zed/discussions/24260 I mentioned
issue.
Make a complex data table example to test the text rendering
performance.
This example also can be an example to show how to build a large data
table.
```bash
cargo run -p gpui --example data_table
```
<img width="2004" alt="image"
src="https://github.com/user-attachments/assets/653771e5-ef08-4d76-97b9-90ea4b78be59"
/>
----
I will try to do some test.
For example: With a threshold for the hold number of caches in
`FrameCache`, and only when the threshold is greater than a certain
number, some caches are released, or when a certain time has passed. I
am not sure if this is feasible.
This example is added to help us to test.
Closes#20762
Release Notes:
- N/A
---
Enable MSAA for Anti-Aliasing to Path (`cx.paint_path`) for drawing a
better vector graphics.
```bash
cargo run -p gpui --example gradient --features macos-blade
cargo run -p gpui --example gradient
cargo run -p gpui --example painting --features macos-blade
cargo run -p gpui --example painting
```
**Before**
<img width="1089" alt="image"
src="https://github.com/user-attachments/assets/0ae7240f-4ba9-4ef5-896c-e436c1282770"
/>
**After**
<img width="944" alt="image"
src="https://github.com/user-attachments/assets/71a07ae8-be54-452c-aacc-b8cec1f810c0"
/>
## TODO
- [x] Support Metal and Blade.
- [x] Detect system support to set up sample count.
- [x] Fix extra lines between Path vertices wait #22808 to merge.
Ref https://github.com/kvark/blade/pull/213
Ask @kvark to review.
I am not sure if there is anything I missed. I modified it according to
the
[particle](https://github.com/kvark/blade/tree/main/examples/particle)
example of Blade project. But the difference is that after the first
MSAA render, I did not do it a second time, I tested it and found it was
not necessary.
Release Notes:
- N/A
---
Split from #21438, this change for make sure the `single_line` mode
Label will not be wrap.
---------
Co-authored-by: Marshall Bowers <elliott.codes@gmail.com>
Release Notes:
- N/A
---
```
cargo run -p gpui --example painting
```
I added this demo to verify the detailed support of Path drawing in
GPUI.
Because of, when we actually used GPUI to draw a 2D line chart, we found
that the straight line Path#line_to did not support `anti-aliasing`, and
the drawn line looked very bad.
As shown in the demo image, if we zoom in on the image, we can clearly
see that all the lines are jagged.
I read and tried to make some appropriate adjustments to the functions
in Path, but since I have no experience in the graphics field, I still
cannot achieve anti-aliasing support so far.
I don't know if I used it wrong somewhere. I checked `curve_to` and
found that the curves drawn have anti-aliasing effects, as shown in the
arc part of the figure below.
<img width="1136" alt="image"
src="https://github.com/user-attachments/assets/4dfb7603-e746-43e9-b737-cff56b56329f">
Release Notes:
- N/A
---
We may only want to set the height of an image to limit the size and
make the width adaptive.
In HTML, we will only set width or height, and the other side will adapt
and maintain the original image ratio.
I changed this because I had a logo image that only to be limited in
height, and then I found that setting the height of the `img` alone
would not display correctly.
I also tried to set `ObjectFit` in this Demo, but it seems that none of
them can achieve the same effect as "After".
## Before
<img width="809" alt="before 2024-09-18 164029"
src="https://github.com/user-attachments/assets/7ba559ed-e53b-43e6-a072-93c8ba5b14ee">
## After
<img width="749" alt="after 2024-09-18 172003"
src="https://github.com/user-attachments/assets/51ee2eba-76b3-400a-abbf-de0e9c4021e2">
Release Notes:
- N/A
---
Add this for let GPUI element to support fade in-out animation.
## Platform test
- [x] macOS
- [x] blade `cargo run -p gpui --example opacity --features macos-blade`
## Usage
```rs
div()
.opacity(0.5)
.bg(gpui::black())
.text_color(gpui::black())
.child("Hello world")
```
This will apply the `opacity` it self and all children to use `opacity`
value to render colors.
## Example
```
cargo run -p gpui --example opacity
cargo run -p gpui --example opacity --features macos-blade
```
<img width="612" alt="image"
src="https://github.com/user-attachments/assets/f1da87ed-31f5-4b55-a023-39e8ee1ba349">
Release Notes:
- N/A
Currently, the `img` element provided by GPUI only supports FilePath or
URL, but in actual applications we need to let `img` load an image
embedded in Assets.
The `svg` element can currently support this, but `img` cannot.
For example:
We have such an Assets directory:
```
assets
|- icons
|- images
|--- foo.png
```
```rs
// If give a path, considered an Asset
img("images/foo.png");
// If give a URI, considered a Remote image
img("https://foo.bar/images/foo.png");
// If give a PathBuf, considered a Local file
img(PathBuf::from("path/to/foo.png"));
```
## Example test
```
cargo run -p gpui --example image
```
<img width="827" alt="image"
src="https://github.com/user-attachments/assets/e45dcf7f-4626-4fb0-aca9-9b6e1045a952">
---------
Co-authored-by: Marshall Bowers <elliott.codes@gmail.com>
Release Notes:
- Fixed default font_family crash on Windows, use `Segoe UI`.
## Crash error message
```
thread 'main' panicked at crates\gpui\src\text_system.rs:150:9:
failed to resolve font 'Helvetica' or any of the fallbacks:
Zed Plex Mono, Helvetica, Cantarell, Ubuntu, Noto Sans, DejaVu Sans
```
Release Notes:
- N/A
----
Continue #14044 for Windows
## The problem
The `cx.open_window` method has provided us a `window_kind` option to
allows creating a Popup kind. This behavior can work on macOS, the popup
kind window have no-border, no-shadow, no-resize, and followed the
`is_movable` if present true it can't move.
This PR to fix those supports on Windows.
The border and shadow still exist, I have tried to use WS_POPUP
window_style, but it will crash:
> This is looks like complex, it is out of my known.
```
Blocking waiting for file lock on build directory
Compiling gpui v0.1.0 (F:\work\zed\crates\gpui)
Finished `dev` profile [unoptimized + debuginfo] target(s) in 13.96s
Running `target\debug\examples\window_positioning.exe`
thread 'main' panicked at F:\Users\jason\.cargo\git\checkouts\blade-b2bcd1de1cf7ab6a\21a56f7\blade-graphics\src\vulkan\init.rs:864:18:
called `Result::unwrap()` on an `Err` value: ERROR_OUT_OF_DEVICE_MEMORY
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
error: process didn't exit successfully: `target\debug\examples\window_positioning.exe` (exit code: 0xc0000409, STATUS_STACK_BUFFER_OVERRUN)
```
So I just make a simple change, to use `WS_EX_TOOLWINDOW` this can
disable resize, and connect `is_movable` to `handle_hit_test_msg` to
disable move, and also no Status Bar icon.
## Before
https://github.com/zed-industries/zed/assets/5518/76740a71-e0ba-401f-958d-f4afdeb417c6
## After
https://github.com/zed-industries/zed/assets/5518/dca49f13-914c-425a-b8b6-b9fc15f8d208
Release Notes:
- Fixed line wrap for CJK characters.
## Demo
https://github.com/zed-industries/zed/assets/5518/c6695bb4-b170-4ce0-9a84-c36b051de438
Fix issues: #4623#11202
### Render case
```
## fr
Bien démarrer avec la documentation GitHub Découvrez comment commencer à créer, à livrer et à gérer des logiciels avec GitHub. Explorez nos produits, inscrivez-vous pour obtenir un compte et connectez-vous à la plus grande communauté de développement du monde.
## zh
GitHub 入门文档 了解如何开始构建、运输和维护具有 GitHub 的软件。 了解我们的产品，注册一个帐户，与世界上最大的发展社区建立联系。
## es
Documentación sobre la introducción a GitHub Aprende cómo comenzar a crear, enviar y mantener software con GitHub. Explora nuestros productos, regístrate para una cuenta y conéctate con la comunidad de desarrollo más grande del mundo.
## kr
GitHub 설명서 시작 GitHub를 사용하여 소프트웨어 빌드, 납품 및 유지 관리를 시작하는 방법을 알아봅니다. 제품을 탐색하고, 계정에 등록하고, 세계 최대의 개발 커뮤니티와 연결합니다.
## ja
GitHub の概要に関するドキュメント GitHub を使用してソフトウェアの構築、出荷、および保守を始める方法を学びます。 当社の製品を探索し、アカウントにサインアップして、世界最大の開発コミュニティと繋がりましょう。
## pt
Documentação de introdução ao GitHub Aprenda a começar a criar, enviar e manter um software com a GitHub. Explore nossos produtos, inscreva-se em uma conta e conecte-se com a maior comunidade de desenvolvimento do mundo.
## ru
Начало работы с документацией по GitHub Узнайте, как начать создание, доставку и обслуживание программного обеспечения с помощью GitHub. Изучите наши продукты, зарегистрируйте учетную запись и присоединитесь к крупнейшему в мире сообществу разработчиков.
```
Release Notes:
- Improved App Menu, add `...` for modal action menu, and group menu
items by type.
In macOS and Windows, the `...` suffix of menu item, is means that will
open a dialog.
https://github.com/zed-industries/zed/assets/5518/8bbd13a7-9144-48b0-9bc8-6651725476f8
Closes https://github.com/zed-industries/zed/issues/8651
Reworks `recent_projects::OpenRecent` action with collab projects in mind:
* keep the "open in new window" behavior for corresponding menu and command entries
* use new, "reuse current window" behavior in the recent projects picker up in the toolbar
This way, old Zed behavior is not customizable, kept as original in all main use cases — so that projects shared via remote entities: a channel and a call, are never accidentally closed, breaking the sharing.
Release Notes:
- Return "open in new window" as default in recent projects
Jason LeeandDanilo Leal
