Closes#22655
A more detailed write-up for this change can be found in the issue
itself. Here, I'm just providing previews after the change. The preview
before the change can be found in the attached issue.
1. While selecting multiple entries, the last clicked entry should be
selected.
[a.webm](https://github.com/user-attachments/assets/2add69c3-82a9-4e45-92e8-366aaf9b298a)
2. When holding `Ctrl`/`Cmd` on an entry, there should be clear visual
feedback to indicate whether the entry is selected or marked.
[b.webm](https://github.com/user-attachments/assets/2cefb8aa-e7d0-4929-9efa-89a4329f428b)
3. When only one entry is marked, but it’s different from the selection,
operations should prioritize the selected entry. This let's you do quick
one-off actions without disrupting the marked state.
[c.webm](https://github.com/user-attachments/assets/8e7ae0c0-4387-49b9-9761-5d02a1c21a84)
4. When more than one entries are marked, operations should prioritize
the marked entries. If the selection differs from the marked entries, it
should not interfere with operations on the marked entries. This let's
you do actions on multiple marked entries without needing to adjust the
selection.
[d.webm](https://github.com/user-attachments/assets/165a74be-cbe9-48ac-b558-2562485ea224)
Release Notes:
- Improved project panel selection, copying, and deletion behavior, to
be more predictable.
Closes#22399
Currently, the target file is being trashed when trashing a symlink, and
the symlink remains intact. Symlinks are not handled separately yet, so
when `open` is used on a symlink, it gets resolved to the target file.
To fix this, we can get the file descriptor of the symlink by passing
`libc::O_PATH | libc::O_NOFOLLOW` flags to `open`, and then pass this
file descriptor to the existing `trash::trash_file` from `ashpd`.
However, this would result in an error because `ashpd` currently does
not support trashing symlink files. I have created an issue for it here:
[https://github.com/bilelmoussaoui/ashpd/issues/255](https://github.com/bilelmoussaoui/ashpd/issues/255).
For the time being, this PR partially fixes the issue by removing the
symlink without trashing so that the target file won't be affected. Once
the upstream bug is fixed, we can switch this remove action back to
trashing.
Release Notes:
- Fixed target file from being trashed when trashing symlink on Linux.
No issue, as the functionality is currently not being used in Zed. This
is more of a GPUI improvement.
Currently, `keyboard_layout` and `on_keyboard_layout_change` are already
handled on macOS. This PR implements the same for X11 and Wayland.
Linux supports up to 4 keyboard layout groups (e.g., Group 0: English
US, Group 1: Bulgarian, etc). On X11 and Wayland, `event` provides a new
active group, which maps to the `layout_index`. We already store keymap
state from where we can get the current `layout_index`. By comparing
them, we determine if the layout has changed.
X11:
<img
src="https://github.com/user-attachments/assets/b528db77-1ff2-4f17-aac5-7654837edeb9"
alt="x11" width="300px" />
Wayland:
<img
src="https://github.com/user-attachments/assets/2b4e2a30-b0f4-495c-96bb-7bca41365d56"
alt="wayland" width="300px" />
Release Notes:
- N/A
This is harmful for user experience and at best requires a user setting.
This was committed as part of
https://github.com/zed-industries/zed/pull/21739 however that change had
no relevant release notes and no relevant settings.
The issue #22343 shows how this can result in user experience
regression: deleting a text fragment can reinsert it back, and it's thus
unclear if the deletion has even worked. Maybe this can be reenabled in
some very restrictive setup, and put behind a setting, but it can't be
unconditional. Completions should activate when the user signals intent
of entering code - for example, if instead of `de` to delete a fragment,
I press `ce` to replace it, I would naturally expect inline completions
to show up.
Note: The linked PR added more code in vim crate to refresh inline
completions in normal mode. I'm keeping that code around in this commit,
so that this can be the minimal fix to the linked issue -- with the
assumption that maybe there's some way in the future to reenable this in
a subset of cases that don't result in confusing / broken UX. If that is
not true the code might need further cleanup. Let me know if you'd
rather see removal of those changes in this PR as well.
Closes#22343.
Release Notes:
- Fixes inline completions showing up in Vim normal mode.
The code we have assumes that when you run commands over ssh they run
in your home directory. This was not true in some cases, and broke SSH
remoting if you had `upload_binary_over_ssh` set.
To reproduce this use Coder and set the `dir` parameter.
Release Notes:
- Fixed SSH remoting in the case that ssh defaults to a non-$HOME
directory.
This PR adds some missing calls to clear the sub-collections in the
`ThreadStore` when we call `ThreadStore::drain` or `ThreadStore::clear`.
Release Notes:
- N/A
The existing code was causing us to constantly re-scan files when
anything changed in the project. Temporarily revert this as we're about
to rework this entire UI with the new primitives.
follow up to https://github.com/zed-industries/zed/pull/22329
Release Notes:
- N/A
---------
Co-authored-by: Cole <cole@zed.dev>
Based on the python https://github.com/zed-industries/zed/issues/21452
and PR https://github.com/zed-industries/zed/pull/22587
I found the same problem with go on windows.
Describe the bug / provide steps to reproduce it
Language server error: gopls
The system cannot find the file specified. (os error 2)
-- stderr--
[ERROR project::lsp_store] Failed to start language server "gopls": The
system cannot find the file specified. (os error 2)
[ERROR project::lsp_store] server stderr: ""
Environment
Windows 11
Go
Release Notes:
- Windows: Fixed `gopls` path construction on Windows 11.
---------
Co-authored-by: Marshall Bowers <elliott.codes@gmail.com>
Release Notes:
- Added support for Google's Gemini 2.0 Flash experimental model.
Note:
Weirdly enough the model is slow on small talk responses like 'hi' (in
my tests) but very fast on things that need more tokens like 'write me a
snake game in python'. Likely an API problem.
TESTED ONLY ON WINDOWS! Would test further but don't have Linux
installed and don't have an Mac. Will likely work everywhere.
Why?:
I think Gemini 2.0 Flash is incredibly good model at coding and
following instructions. I think it would be nice to have it in the
editor. I did as minimal changes as possible while adding the model and
streaming validation. I think it's worth merging the commits as they
bring good improvements.
---------
Co-authored-by: Marshall Bowers <elliott.codes@gmail.com>
This will allow the diagnostic popover to be displayed even if hovering
an invisible char.
Beyond that, it solves a rare `DiagnosticPopover` corner case:
* Supports moving the selection to a diagnostic when `GoToDiagnostic` is
done while hovering, based on `group_id`
* Provides Diagnostic values with `group_id: 0` providing information on
hover about invisible characters.
So, `GoToDiagnostic` would navigate to the very first error produced by
a language server. Really not a big deal of course.
Release Notes:
- N/A
Closes#22433
Before/After (macOS):
For some reason the non-blurred one seems much lower quality, so we may
need to tinker with the samples, or something else.
I'm unsure if this is a problem on Linux/in the Blade renderer, but
since no changes were made outside of the medal shaders we can probably
take this macOS-specific win for now.
Release Notes:
- gpui: Fixed an issue where shadows with a `blur_radius` of 0 would not
render.
This PR adds timeouts when resolving feature flags during workspace
panel initialization so that we don't block indefinitely if Zed is not
connected to the internet.
Right now we wait for 5 seconds, but this value was chosen arbitrarily.
Release Notes:
- N/A
Namely, just removing the unnecessary extra line dividers and adding a
super subtle background color to the diagnostic message to create a bit
of separation/hierarchy.
<img width="800" alt="Screenshot 2025-01-04 at 9 46 03 PM"
src="https://github.com/user-attachments/assets/d62883b9-ed76-4fbb-b9c1-b55146eaeed4"
/>
Release Notes:
- N/A
This reverts commit 3ae6aa0e4d.
If "group_id = 0" really did mean a diagnostic not from a language
server then various methods related to diagnostic set would need to be
updated. Something like [this
diff](https://gist.github.com/mgsloan/e902153bcaec207b39260a8f40d3134d).
Plan instead is to use InfoPopover instead of DiagnosticPopover for
these.
Release Notes:
- N/A
No need to find or store the primary range ahead of time as it's found
by `activate_diagnostics`.
Not entirely sure we should still even have the special case when the
popover is visible. It does support the keyboard interaction of opening
hover followed by jumping to the primary position, but that seems pretty
undiscoverable.
Support for clicking the hover to navigate to the primary diagnostic was
removed in https://github.com/zed-industries/zed/pull/3408
Release Notes:
- N/A
Should resolve https://github.com/zed-industries/zed/issues/21714.
In some conditions that I'm not sure of, Zed sends LSP requests with
`start > end` position, and zls has an [assertion for end >=
start](f253553b82/src/offsets.zig (L492)),
and that causes zls to crash, like:
```bash
# first `textDocument/inlayHint` request with `end >= start`
[2025-01-05T19:33:09+00:00 TRACE lsp] outgoing message:{"jsonrpc":"2.0","id":1043,"method":"textDocument/inlayHint","params":{"textDocument":{"uri":"file:///Users/burak/Code/parzig/src/parquet/decoding.zig"},"range":{"start":{"line":0,"character":0},"end":{"line":24,"character":0}}}}
# successful response
[2025-01-05T19:33:09+00:00 TRACE lsp::input_handler] incoming message: {"jsonrpc":"2.0","id":1043,"result":[{"position":{"line":0,"character":9},"label":": type","kind":1,"paddingLeft":false,"paddingRight":false},{"position":{"line":1,"character":22},"label":": type","kind":1,"paddingLeft":false,"paddingRight":false},{"position":{"line":4,"character":13},"label":": [](unknown type)","kind":1,"paddingLeft":false,"paddingRight":false},{"position":{"line":4,"character":30},"label":"T:","kind":2,"tooltip":{"kind":"markdown","value":"```zig\ncomptime type\n```"},"paddingLeft":false,"paddingRight":true},{"position":{"line":4,"character":33},"label":"n:","kind":2,"tooltip":{"kind":"markdown","value":"```zig\nusize\n```"},"paddingLeft":false,"paddingRight":true},{"position":{"line":5,"character":23},"label":": bool","kind":1,"paddingLeft":false,"paddingRight":false},{"position":{"line":6,"character":19},"label":": usize","kind":1,"paddingLeft":false,"paddingRight":false},{"position":{"line":9,"character":26},"label":": [](unknown type)","kind":1,"paddingLeft":false,"paddingRight":false},{"position":{"line":9,"character":43},"label":"T:","kind":2,"tooltip":{"kind":"markdown","value":"```zig\ncomptime type\n```"},"paddingLeft":false,"paddingRight":true},{"position":{"line":9,"character":47},"label":"n:","kind":2,"tooltip":{"kind":"markdown","value":"```zig\nusize\n```"},"paddingLeft":false,"paddingRight":true},{"position":{"line":21,"character":13},"label":": [](unknown type)","kind":1,"paddingLeft":false,"paddingRight":false},{"position":{"line":21,"character":30},"label":"T:","kind":2,"tooltip":{"kind":"markdown","value":"```zig\ncomptime type\n```"},"paddingLeft":false,"paddingRight":true},{"position":{"line":21,"character":33},"label":"n:","kind":2,"tooltip":{"kind":"markdown","value":"```zig\nusize\n```"},"paddingLeft":false,"paddingRight":true},{"position":{"line":22,"character":33},"label":"T:","kind":2,"tooltip":{"kind":"markdown","value":"```zig\ncomptime type\n```"},"paddingLeft":false,"paddingRight":true},{"position":{"line":22,"character":36},"label":"buf:","kind":2,"tooltip":{"kind":"markdown","value":"```zig\n[]T\n```"},"paddingLeft":false,"paddingRight":true},{"position":{"line":22,"character":41},"label":"bit_width:","kind":2,"tooltip":{"kind":"markdown","value":"```zig\nu8\n```"},"paddingLeft":false,"paddingRight":true},{"position":{"line":22,"character":52},"label":"reader:","kind":2,"tooltip":{"kind":"markdown","value":"```zig\nanytype\n```"},"paddingLeft":false,"paddingRight":true}]}
[2025-01-05T19:33:09+00:00 TRACE lsp] Took 14.855ms to receive response to "textDocument/inlayHint" id 1043
# problematic `textDocument/inlayHint` request with `start > end`
[2025-01-05T19:33:09+00:00 TRACE lsp] outgoing message:{"jsonrpc":"2.0","id":1044,"method":"textDocument/inlayHint","params":{"textDocument":{"uri":"file:///Users/burak/Code/parzig/src/parquet/decoding.zig"},"range":{"start":{"line":50,"character":25},"end":{"line":25,"character":0}}}}
# zls crashes here, and after this point, all LSP requests fail
[2025-01-05T19:33:09+00:00 TRACE lsp] incoming stderr message:thread 5391652 panic: reached unreachable code
[2025-01-05T19:33:09+00:00 ERROR lsp] cannot read LSP message headers
```
In LSP specification for
[`Range`](https://microsoft.github.io/language-server-protocol/specifications/lsp/3.17/specification/#range)
type, it says:
> ... If you want to specify a range that contains a line including the
line ending character(s) then use an end position denoting the start of
the next line.
I feel like zls's assertion is sensible, so I've updated the generic
`range_to_lsp` function rather than doing something specific to zls. But
let me know if this seems incorrect.
zls was crashing after 5-10 minutes of working with a Zig codebase
before, and after this change, I tested for an hour and didn't
experience any crashes.
Release Notes:
- Ensure `end >= start` in `lsp::Range`, which should fix Zig/zls
crashes.
---------
Co-authored-by: Michael Sloan <michael@zed.dev>
In particular, `DiagnosticPopover` both:
* Supports moving the selection to a diagnostic when clicked, based on
`group_id`
* Provides Diagnostic values with `group_id: 0` providing informztion on
hover about invisible characters.
So, clicking such a popover would navigate to the very first error
produced by a language server. Really not a big deal of course, but
seems good to fix as it might result in surprising behavior in other
future circumstances
Release Notes:
- N/A
In #22620 `diagnostic_group` was modified to return results for
multibuffers, but was returning singleton buffer points. `hover_popover`
uses it to find the jump target for clicking the popup - which doesn't
seem to be working right now but that's a separate issue. Now that
`diagnostic_group` is returning values in multibuffers converting these
to anchors was crashing.
Also resolves a potential bug - if folding in multibuffers was supported
then "Go To Diagnostics" would not properly skip diagnostics from folded
regions.
Release Notes:
- N/A
Closes#15705
When opening a file from the terminal, if the file path is relative, we
attempt to guess all possible paths where the file could be. This
involves generating paths for each worktree, the current terminal
directory, etc. For example, if we have two worktrees, `dotfiles` and
`example`, and `foo.txt` in `example/a`, the generated paths might look
like this:
- `/home/tims/dotfiles/../example/a/foo.txt` from the `dotfiles`
worktree
- `/home/tims/example/../example/a/foo.txt` from the `example` worktree
- `/home/tims/example/a/foo.txt` from the current terminal directory
(This is already canonicalized)
Note that there should only be a single path, but multiple paths are
created due to missing canonicalization.
Later, when opening these paths, the worktree prefix is stripped, and
the remaining path is used to open the file in its respective worktree.
As a result, the above three paths would resolve like this:
- `../example/a/foo.txt` as the filename in the `dotfiles` worktree
(Ghost file)
- `../example/a/foo.txt` as the filename in the `example` worktree
(Ghost file)
- `foo.txt` as the filename in the `a` directory of the `example`
worktree (This opens the file)
This PR fixes the issue by canonicalizing these paths before adding them
to the HashSet.
Before:
After:
Release Notes:
- Fixed ghost files appearing in the project panel when clicking
relative paths in the terminal.
Closes https://github.com/zed-industries/zed/issues/22642
In Zed, Rust's label generators expected the details to come in ` (use
std.foo.Bar)` form, but recently, r-a started to send these details
without the leading whitespace which broke the code generation.
The PR makes LSP results parsing more lenient to work with both details'
forms.
Release Notes:
- Fixed Rust completion labels not showing the imports
- [x] Rewrite worktree git handling
- [x] Fix tests
- [x] Fix `test_propagate_statuses_for_repos_under_project`
- [x] Replace `WorkDirectoryEntry` with `WorkDirectory` in
`RepositoryEntry`
- [x] Add a worktree event for capturing git status changes
- [x] Confirm that the local repositories are correctly updating the new
WorkDirectory field
- [x] Implement the git statuses query as a join when pulling entries
out of worktree
- [x] Use this new join to implement the project panel and outline
panel.
- [x] Synchronize git statuses over the wire for collab and remote dev
(use the existing `worktree_repository_statuses` table, adjust as
needed)
- [x] Only send changed statuses to collab
Release Notes:
- N/A
---------
Co-authored-by: Cole Miller <cole@zed.dev>
Co-authored-by: Mikayla <mikayla@zed.com>
Co-authored-by: Nathan <nathan@zed.dev>
Closes#22607
Symlinks can be absolute or relative. When using
[stow](https://www.gnu.org/software/stow/) to manage dotfiles, it
creates relative symlinks to the target files.
For example:
- Original file: `/home/tims/dotfiles/zed/setting.json`
- Symlink path: `/home/tims/.config/zed/setting.json`
- Target path (relative to symlink): `../../dotfiles/zed/setting.json`
The issue is that you can’t watch the symlink path because it’s relative
and doesn't include the base path it is relative to. This PR fixes that
by converting relative symlink paths to absolute paths.
- Absolute path (after parent join):
`/home/tims/.config/zed/../../dotfiles/zed/setting.json` (This works)
- Canonicalized path (from absolute path):
`/home/tims/dotfiles/zed/setting.json` (This works too, just more
cleaner)
Release Notes:
- Fix issue where items on the Welcome page could not be toggled on
Linux when using Stow to manage dotfiles
Closes#22326
This PR adds process PID information to window created by X11, so that
window manager can identify which process this window belongs to.
Without this property, the window manager would have no reliable way to
know which process created this window.
In original issue, `robotgo` throws error on `x, y, w, h :=
robotgo.GetBounds(pid)` this method. If we go deeper into the source
code of `robotgo`, it calls `GetXidFromPid` which goes through all
windows, and tries to check for provided pid. Hence, when it tries to do
that for Zed, it fails and returns `0, err` to caller.
```go
// Robotgo source code trying to look through all windows and query pid
// GetXidFromPid get the xid from pid
func GetXidFromPid(xu *xgbutil.XUtil, pid int) (xproto.Window, error) {
windows, err := ewmh.ClientListGet(xu)
if err != nil {
return 0, err
}
for _, window := range windows {
wmPid, err := ewmh.WmPidGet(xu, window)
if err != nil {
return 0, err
}
if uint(pid) == wmPid {
return window, nil
}
}
return 0, errors.New("failed to find a window with a matching pid.")
}
```
Querying for pid for active Zed window:
Before:
```sh
tims@lemon ~/w/go-repro [127]> xprop -root _NET_ACTIVE_WINDOW
_NET_ACTIVE_WINDOW(WINDOW): window id # 0x4e00002
tims@lemon ~/w/go-repro> xprop -id 0x4e00002 _NET_WM_PID
_NET_WM_PID: not found.
```
After:
```sh
tims@lemon ~/w/go-repro> xprop -root _NET_ACTIVE_WINDOW
_NET_ACTIVE_WINDOW(WINDOW): window id # 0x4e00002
tims@lemon ~/w/go-repro> xprop -id 0x4e00002 _NET_WM_PID
_NET_WM_PID(CARDINAL) = 103548
tims@lemon ~/w/go-repro>
```
Correct zed process PID (below) assosiated with zed window (shown
above):
Release Notes:
- Fix `robotgo` failing when Zed window is open on Linux
Closes#22578
Currently, the `hovered` boolean in the window state is only updated by
the `WM_MOUSELEAVE` event, which fires when the mouse cursor leaves the
window's working area. This means that when the user moves the cursor
from the window to the title bar, `hovered` is set to `false`. Later in
the code, this flag is used to determine the cursor style and check if
the cursor is over the correct window.
The `hovered` boolean should remain active even when the mouse is over
non-client items, such as the title bar or window borders. This PR fixes
that by using `WM_NCMOUSELEAVE` event, which is triggered when the mouse
leaves non-client items. This event is used to update the `hovered`
boolean accordingly.
Now, `hovered` is `true` when the mouse is over the window's working
area, as well as non-client areas like the title bar.
More context:
- Existing: `dwFlags: TME_LEAVE` tracks window area mouse leaves, which
is used in `handle_mouse_move_msg` func.
- New: `dwFlags: TME_LEAVE | TME_NONCLIENT` tracks non-client mouse
leaves, which is used in `handle_nc_mouse_move_msg` func.
Preview:
https://github.com/user-attachments/assets/b319303f-81b9-45cb-bf0c-535a59b96561
Release Notes:
- Fix cursor style not changing on hover over items in the title bar on
Windows
Release Notes:
- Added the ability to specify additional beta headers for custom
Anthropic models.
---------
Co-authored-by: David Soria Parra <167242713+dsp-ant@users.noreply.github.com>
Co-authored-by: Marshall Bowers <elliott.codes@gmail.com>
Closes #ISSUE
- #21452
Describe the bug / provide steps to reproduce it
Language server error: pylsp
failed to spawn command. path:
"C:\Users\AppData\Local\Zed\languages\pylsp\pylsp-venv\bin\pylsp",
working directory: "D:\Coding\Python", args: []
-- stderr--
Environment
- Windows 11
- python
Release Notes:
- Windows: Fixed the path building used to run `pip` commands in the
venv generated on Windows 11.
---------
Co-authored-by: Marshall Bowers <elliott.codes@gmail.com>
Removed a settings update that should have been removed in the 0.148.0
release.
I am not sure if there is a tracking issue, but I identified this check
for outdated settings that should not be needed anymore. I investigated
a bit and did not find any conflicts or UB as a result of removing this
code.
Release Notes:
- N/A
---------
Co-authored-by: Marshall Bowers <elliott.codes@gmail.com>
This PR fixes an issue introduced in #21939 where the list of models in
the language model selector could be outdated.
Since we're no longer recreating the picker each render, we now need to
make sure we are updating the list of models accordingly when there are
changes to the language model providers.
I noticed it specifically in Assistant1.
Release Notes:
- Fixed a staleness issue with the language model selector.
