Additionally, mark context menu entry as disabled when the action would
fail (untitled buffer, collab sessions).
Supersedes #18584
Release Notes:
- Fixed "Reveal in Finder/File Manager", "Copy Path", "Copy Relative
Path" and "Copy file location" actions not working with multibuffers.
We introduced a memory leak in #18572, which meant that `Drop` was never
called on `SshRemoteConnection`, meaning that the ssh process kept
running
Co-Authored-by: Thorsten <thorsten@zed.dev>
Release Notes:
- N/A
---------
Co-authored-by: Thorsten <thorsten@zed.dev>
Also, fix visual bug when pressing escape with a non-empty selection in
a deleted text block.
Release Notes:
- N/A
Co-authored-by: Antonio <antonio@zed.dev>
Closes#18531
This PR limits the range of values that can be set for `FontWeight`.
Since any value less than 1.0 or greater than 999.9 causes Zed to crash
on Windows, I’ve restricted `FontWeight` to this range.
I could apply this constraint only on Windows, but considering the
documentation at https://zed.dev/docs/configuring-zed#buffer-font-weight
indicates that `FontWeight` should be between 100 and 900, I thought it
might be a good idea to apply this restriction in the settings.
Release Notes:
- Changed `ui_font_weight` and `buffer_font_weight` settings to require
values to be between `100` and `950` (inclusive).
---------
Co-authored-by: Marshall Bowers <elliott.codes@gmail.com>
This PR makes it so the proposed changes editor works with the workspace
navigation history.
This allows for easily navigating back to the proposed changes editor
after opening one of the excerpts into the base buffer.
Release Notes:
- N/A
This PR reverts the changes introduced via #18164. As shown in the video
below, once you `hide` the app, there is essentially no way to bring it
back. I must emphasize that the window logic on Windows is entirely
different from macOS. On macOS, when you `hide` an app, its icon always
remains visible in the dock, and you can always bring the hidden app
back by clicking that icon. However, on Windows, there is no such
mechanism—the app is literally hidden.
I think the `hide` feature should be macOS-only.
https://github.com/user-attachments/assets/65c8a007-eedb-4444-9499-787b50f2d1e9
Release Notes:
- N/A
This restructures the code in `remote` so that it's easier to replace
the current SSH connection with a new one in case of
disconnects/reconnects.
Right now, it successfully reconnects, BUT we're still missing the big
piece on the server-side: keeping the server process alive and
reconnecting to the same process that keeps the project-state.
Release Notes:
- N/A
---------
Co-authored-by: Bennet <bennet@zed.dev>
Closes #14089, #14416, #15970, #17230, #18485
Release Notes:
- Fixed some cases where Linux X11 mouse scrolling doesn't work at all
(#14089, ##15970, #17230)
- Fixed handling of switching between Linux X11 devices used for
scrolling (#14416, #18485)
Change details:
Also includes the commit from PR #18317 so I don't have to deal with
merge conflicts.
* Now uses valuator info from slave pointers rather than master. This
hopefully fixes remaining cases where scrolling is fully
broken. https://github.com/zed-industries/zed/issues/14089,
https://github.com/zed-industries/zed/issues/15970,
https://github.com/zed-industries/zed/issues/17230
* Per-device recording of "last scroll position" used to calculate
deltas. This meant that swithing scroll devices would cause a sudden
jump of scroll position, often to the beginning or end of the
file (https://github.com/zed-industries/zed/issues/14416).
* Re-queries device metadata when devices change, so that newly
plugged in devices will work, and re-use of device-ids don't use old
metadata with a new device.
* xinput 2 documentation describes support for multiple master
devices. I believe this implementation will support that, since now it
just uses `DeviceInfo` from slave devices. The concept of master
devices is only used in registering for events.
* Uses popcount+bit masking to resolve axis indexes, instead of
iterating bit indices.
---------
Co-authored-by: Thorsten Ball <mrnugget@gmail.com>
Closes#17605
Watches for target paths if file watched is a symlink in Linux.
This will check if the generated `notify::Event` has any paths matching
the `root_path` and if the file is a symlink it will also check if the
path matches the `target_root_path` (the path that the symlink is
pointing to)
Release Notes:
- Added file watching for symlinks
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">
**Clipboard Behavior on Windows Under This PR:**
| User Action | Zed’s Behavior |
| ------------------- |
-------------------------------------------------- |
| Paste PNG | Worked |
| Paste JPEG | Worked |
| Paste WebP | Worked, but not in the way you expect (see Issue section
below) |
| Paste GIF | Partially worked (see Issue section below) |
| Paste SVG | Partially worked (see Issue section below) |
| Paste BMP | Worked, but not in the way you expect (see Issue section
below) |
| Paste TIFF | Worked, but not in the way you expect (see Issue section
below) |
| Paste Files | Worked, same behavior as macOS |
| Copy image in Zed | Not tested, as I couldn’t find a way to copy
images |
---
**Differences Between the Windows and macOS Clipboard**
The clipboard functionality on Windows differs significantly from macOS.
On macOS, there can be multiple items in the clipboard, whereas, on
Windows, the clipboard holds only a single item. You can retrieve
different formats from the clipboard, but they are all just different
representations of the same item.
For example, when you copy a JPG image from Microsoft Word, the
clipboard will contain data in several formats:
- Microsoft Office proprietary data
- JPG format data
- PNG format data
- SVG format data
Please note that these formats all represent the same image, just in
different formats. This is due to compatibility concerns on Windows, as
various applications support different formats. Ideally, multiple
formats should be placed on the clipboard to support more software.
However, in general, supporting PNG will cover 99% of software, like
Chrome, which only supports PNG and BMP formats.
Additionally, since the clipboard on Windows only contains a single
item, special handling is required when copying multiple objects, such
as text and images. For instance, if you copy both text and an image
simultaneously in Microsoft Word, Microsoft places the following data on
the clipboard:
- Microsoft Office proprietary data containing a lot of content such as
text fonts, sizes, italics, positioning, image size, content, etc.
- RTF data representing the above content in RTF format
- HTML data representing the content in HTML format
- Plain text data
Therefore, for the current `ClipboardItem` implementation, if there are
multiple `ClipboardEntry` objects to be placed on the clipboard, RTF or
HTML formats are required. This PR does not support this scenario, and
only supports copying or pasting a single item from the clipboard.
---
**Known Issues**
- **WebP, BMP, TIFF**: These formats are not explicitly supported in
this PR. However, as mentioned earlier, in most cases, there are
corresponding PNG format data on the clipboard. This PR retrieves data
via PNG format, so users copying images in these formats from other
sources will still see the images displayed correctly.
- **GIF**: In this PR, GIFs are displayed, but for GIF images with
multiple frames, the image will not animate and will freeze on a single
frame. Since I observed the same behavior on macOS, I believe this is
not an issue with this PR.
- **SVG**: In this PR, only the top-left corner of the SVG image is
displayed. Again, I observed the same behavior on macOS, so I believe
this issue is not specific to this PR.
---
I hope this provides a clearer understanding. Any feedback or
suggestions on how to improve this are welcome.
Release Notes:
- N/A
This is a follow-up to #18530 thanks to this comment here:
https://github.com/zed-industries/zed/pull/18530#issuecomment-2382870564
In short: it fixes the `blinking` setting and the `cursor_shape` setting
as it relates to blinking.
Turns out our `blinking` setting was always the wrong value when using
`terminal_controlled` and the terminal _would_ control the blinking.
Example script to test with:
```bash
echo -e "0 normal \x1b[\x30 q"; sleep 2
echo -e "1 blink block \x1b[\x31 q"; sleep 2
echo -e "2 solid block \x1b[\x32 q"; sleep 2
echo -e "3 blink under \x1b[\x33 q"; sleep 2
echo -e "4 solid under \x1b[\x34 q"; sleep 2
echo -e "5 blink vert \x1b[\x35 q"; sleep 2
echo -e "6 solid vert \x1b[\x36 q"; sleep 2
echo -e "0 normal \x1b[\x30 q"; sleep 2
echo -e "color \x1b]12;#00ff00\x1b\\"; sleep 2
echo -e "reset \x1b]112\x1b\\ \x1b[\x30 q"
```
Before the changes in here, this script would set the cursor shape and
the blinking, but the blinking boolean would always be wrong.
This change here makes sure that it works consistently:
- `terminal.cursor_shape` only controls the *default* shape of the
terminal, not the blinking.
- `terminal.blinking = on` means that it's *always* blinking, regardless
of what terminal programs want
- `terminal.blinking = off` means that it's *never* blinking, regardless
of what terminal programs want
- `terminal.blinking = terminal_controlled (default)` means that it's
blinking depending on what terminal programs want. when a terminal
program resets the cursor to default, it sets it back to
`terminal.cursor_shape` if that is set.
Release Notes:
- Fixed the behavior of `{"terminal": {"blinking":
"[on|off|terminal_controlled]"}` to work correctly and to work correctly
when custom `cursor_shape` is set.
- `terminal.cursor_shape` only controls the *default* shape of the
terminal, not the blinking.
- `terminal.blinking = on` means that it's *always* blinking, regardless
of what terminal programs want
- `terminal.blinking = off` means that it's *never* blinking, regardless
of what terminal programs want
- `terminal.blinking = terminal_controlled (default)` means that it's
blinking depending on what terminal programs want. when a terminal
program resets the cursor to default, it sets it back to
`terminal.cursor_shape` if that is set.
Demo:
https://github.com/user-attachments/assets/b3fbeafd-ad58-41c8-9c07-1f03bc31771f
Co-authored-by: Bennet <bennet@zed.dev>
I've been running with direct direnv loading for a while now and haven't
experienced any significant issues other than #18473. Making it default
would make direnv integration more reliable and consistent. I've also
updated the docs a bit to ensure that they represent current status of
direnv integration
Release Notes:
- Made direnv integration use direct (`direnv export json`) mode by
default instead of relying on a shell hook, improving consistency and
reliability of direnv detection
This fixes something that I felt was off for a while. Previously, when
you'd click on the titlebar to move the window, the titlebar would only
change its background once the moving starts, but not on mouse-down.
That felt really off, since the moving is down with mouse-down and move,
so I think giving the user feedback about the mouse-down event makes
more sense.
I know there's a subjectivity to this change, so I'm ready to hear other
opinions, but for now I want to go with this.
Release Notes:
- N/A
This builds on top of @Yevgen's #15840 and combines it with the settings
names introduced in #17572.
Closes#4731.
Release Notes:
- Added a setting for the terminal's default cursor shape. The setting
is `{"terminal": {"cursor_shape": "block"}}``. Possible values: `block`,
`bar`, `hollow`, `underline`.
Demo:
https://github.com/user-attachments/assets/96ed28c2-c222-436b-80cb-7cd63eeb47dd
Closes#18393
Release Notes:
- Added a `ZED_RC_TOOLKIT_PATH` env variable so `winresource` crate can fetch the RC executable path correctly on some configurations
This is a partial revert of e6c1c51b37, which removed the middle-click
pasting on linux (both x11 & wayland). It also restores the
`middle_click_paste` option behavior which became unexistent.
Release Notes:
- Restore Linux middle-click pasting.
I made a few tree-sitter queries for improving the highlighting of C++.
There is one query that I'm not totally certain about and would
appreciate some feedback on it, the one that concerns attributes.
Many editor only highlight the identifier as a keyword (This is the
behavior implemented in this commit), while others, for example the
tree-sitter plugin for neovim, tags the entire attribute for
highlighting (double brackets included). I don't know which one is
preferable. Here are screenshots of the two versions:
Release Notes:
- Fixed C++ attributes identifiers being wrongly highlighed through the
tag "variable"
- C++ attribute identifiers (nodiscard,deprecated, noreturn, etc.. ) are
now highlighted through the tag "keyword"
- Changed C++ primitives types (void, bool, int, size_t, etc.. ) to no
longer be highlighted with the tag "keyword", they can now be
highlighted by the tag "type.primitive".
- Added a tag "concept" for highlighting C++ concept identifiers. (This
tag name has been chosen to be the same than the one returned by
clangd's semantic tokens)
Allows to split log view, and opens it split on the right, same as the
syntax tree view.
Release Notes:
- Improved language server log panel split ergonomics
Once again aping after what winit does - since we always want to have
the whole window blurred there is apparently no need to specify a blur
region at all. Rounded corners would be the exception, but that is not
possible with the current protocol (it is planned for the vendor-neutral
version though!)
This eliminates the problem where only a fixed region of the window
would get blurred if the window was resized to be larger than at launch.
Also a drive-by comment grammar fix 😉
Release Notes:
- Fixed blur region handling on Plasma/Wayland
Before this we we would stop loading the environment if the call to
direnv failed, which is not necessary in any way
cc @mrnugget
Release Notes:
- Fixed the environment not loading if `direnv` mode is set to `direct`
and `.envrc` is not allowed
If specifying a formatter in the settings like this:
"languages": {
"foo": {
"formatter": {
"external": {
"command": "/path/to/foo-formatter"
}
}
}
}
Zed will show an error like this:
Invalid user settings file
data did not match any variant of untagged enum SingleOrVec
This is because the arguments are not optional. The error is hard to
understand, so let's make the arguments actually optional, which makes
the above settings snippet valid.
Release Notes:
- Make external formatter arguments optional
We are going to use this in the multi-buffer to produce a summary for an
`Excerpt` that contains a `Range<Anchor>`.
Release Notes:
- N/A
Co-authored-by: Nathan <nathan@zed.dev>
We are now using the `view release notes locally` action when clicking
on the update toast - the endpoint for this action does not currently
return anything for valid for these channels, as we don't have support
yet for diffing between these builds, so for now, [continue to do what
the `view release notes` action did and just send the user to the commit
view on
GitHub](caffb2733f/crates/auto_update/src/auto_update.rs (L255-L260)).
It is a bit counterintuitive to send the user to the browser when using
the "local" action, but this is just a patch in the interim.
If we make adjustments to our channels to keep the nightly tag stable
and add some sort of unique suffix, like a timestamp, we can then adjust
things to return these in the request body and show them in the editor.
Release Notes:
- N/A
Closes https://github.com/zed-industries/zed/issues/17527
I think we are ok to switch to using the local action now. There are a
few things we don't support, like media, but we don't include media
directly too often, and I think this might help push the community to
maybe add support for it. That being said, I updated the markdown coming
back from the endpoint to include links to the web version of the
release notes, so they can always hop over to that version, if they
would like.
https://github.com/user-attachments/assets/b4d207a7-1640-48f1-91d0-94537f74116c
All forming of the Markdown happens in the endpoint, so if someone with
a better eye wants to update this, you can do that here:
0e5923e3e7/src/pages/api/release_notes/v2/%5Bchannel_type%5D/%5Bversion%5D.ts (L50-L62)
Release Notes:
- Changed the `view the release notes` button in the update toast to
trigger the local release notes action.
I have just recently discovered this keybinding myself out of talking to
folks, ha. The tooltip here might ease the discovery for other folks in
the future.
<img width="700" alt="Screenshot 2024-09-27 at 11 04 28 PM"
src="https://github.com/user-attachments/assets/844d3b55-15af-47f7-a8db-5c8832ceba29">
---
Release Notes:
- N/A
Fixes https://github.com/zed-industries/zed/issues/18405
In https://github.com/zed-industries/zed/pull/18313, we introduced a
problem where git addition highlights might spuriously return when
undoing certain changes. It turned out, there were already some cases
where git hunk highlighting was incorrect when editing at the boundaries
of expanded diff hunks.
In this PR, I've introduced a test helper method for more rigorously
(and readably) testing the editor's git state. You can assert about the
entire state of an editor's diff decorations using a formatted diff:
```rust
cx.assert_diff_hunks(
r#"
- use some::mod1;
use some::mod2;
const A: u32 = 42;
- const B: u32 = 42;
const C: u32 = 42;
fn main() {
- println!("hello");
+ //println!("hello");
println!("world");
+ //
+ //
}
fn another() {
println!("another");
+ println!("another");
}
- fn another2() {
println!("another2");
}
"#
.unindent(),
);
```
This will assert about the editor's actual row highlights, not just the
editor's internal hunk-tracking state.
I rewrote all of our editor diff tests to use these more high-level
assertions, and it caught the new bug, as well as some pre-existing bugs
in the highlighting of added content.
The problem was how we *remove* highlighted rows. Previously, it relied
on supplying exactly the same range as one that we had previously
highlighted. I've added a `remove_highlighted_rows(ranges)` APIs which
is much simpler - it clears out any row ranges that intersect the given
ranges (which is all that we need for the Git diff use case).
Release Notes:
- N/A
