Reverts zed-industries/zed#41384
The branch-protection rules work much better when there is a Job that
runs every time and can be depended on to pass, we no longer have this.
Release Notes:
- N/A
- Declare UAC support. This will prevent Windows from flagging
`auto_update_helper.exe` as a legacy setup program that needs to run as
administrator.
- Declare support for Windows 10. This will stop Windows from applying
various application compatibility profiles.
The UAC policy is not really appropriate to apply to all GPUI
applications (e.g. an installer written in GPUI may want to declare
itself as `requireAdministrator` instead of `asInvoker`). I tried
splitting this into a Zed.exe-only manifest and enabling manifest file
merging, but I ran out of my time-box. We can fix this later if this is
flagged by GPUI users.
Release Notes:
- N/A
Follow up for: #41304
Splits CI tests (cherry-picks and PRs only for now) into separate
workflows using `gh-workflow`. Includes a couple restructures to
- run more things in parallel
- remove our previous shell script based checking to filter tests based
on files changed, instead using the builtin `paths:` workflow filters
Splitting the docs/style/rust tests & checks into separate workflows
means we lose the complete summary showing all the tests in one view,
but it's possible to re-add in the future if we go back to checking what
files changed ourselves or always run everything.
Release Notes:
- N/A *or* Added/Fixed/Improved ...
---------
Co-authored-by: Conrad <conrad@zed.dev>
Release Notes:
- Remove Windows/Linux from Getting Started
- Consolidate download & system into new Installation page
- Move Remote Dev out of Windows and into Remote Development
- Add Uninstall page
- Add updates page
- Remove addl learning materials from intro
Thread stacks in rust by default have 2 megabytes of stack which for
sumtrees (or ropes in this case) can easily be exceeded depending on the
workload.
Release Notes:
- Fixed stack overflows when constructing large ropes
This PR hides the worktree root name from the path prefix displayed when
you @-mention a file or directory in the agent panel. Given the tight UI
real state we have, I believe that having the project name in there is
redundant—the project you're in is already displayed in the title bar.
Not only it was the very first word you'd see after the file's name, but
it also made the path longer than it needs to. A bit of a design clean
up here :)
(PS: We still show the project name if there are more than one in the
same workspace.)
Release Notes:
- N/A
This PR adds a setting to prevent projects from being shared in public
channels.
This can be enabled by adding the following to the project settings
(`.zed/settings.json`):
```json
{
"prevent_sharing_in_public_channels": true
}
```
This will then disable the "Share" button when not in a private channel:
<img width="380" height="115" alt="Screenshot 2025-10-28 at 2 28 10 PM"
src="https://github.com/user-attachments/assets/6761ac34-c0d5-4451-a443-adf7a1c42bcd"
/>
Release Notes:
- collaboration: Added a `prevent_sharing_in_public_channels` project
setting for preventing projects from being shared in public channels.
Was previously sending all of the tools to the LLM on the first message
of a conversation regardless of the selected agent profile. This added
extra context, and tended to create scenarios where the LLM would
attempt to use the tool and it would fail since it was not available.
To reproduce, create a new conversation where you ask the Minimal mode
which tools it has access to, or try to write to a file.
Release Notes:
- Fixed an issue in the agent where all tools would be presented as
available even when using the `Minimal` profile
---------
Co-authored-by: Bennet Bo Fenner <bennetbo@gmx.de>
Symbol completions in the context picker now show "SymbolName
filename.txt L123" instead of just "SymbolName". This helps distinguish
between symbols with the same name in different files.
## Motivation
I noticed that the message prompt editor only showed symbol names
without any context, making it hard to use in large projects with many
symbols sharing the same name. The inline prompt editor already includes
file context for symbol completions, so I brought that same UX
improvement to the message prompt editor. I've decided to match the
highlighting style with directory completion entries from the message
editor.
## Changes
- Extract file name from both InProject and OutsideProject symbol
locations
- Add `build_symbol_label` helper to format labels with file context
- Update test expectation for new label format
## Screenshots
Inline Prompt Editor
<img width="843" height="334" alt="Screenshot 2025-10-17 at 17 47 52"
src="https://github.com/user-attachments/assets/16752e1a-0b8c-4f58-a0b2-25ae5130f18c"
/>
Old Message Editor:
<img width="466" height="363" alt="Screenshot 2025-10-17 at 17 31 44"
src="https://github.com/user-attachments/assets/900659f3-0632-4dff-bc9a-ab2dad351964"
/>
New Message Editor:
<img width="482" height="359" alt="Screenshot 2025-10-17 at 17 31 23"
src="https://github.com/user-attachments/assets/bf382167-f88b-4f3d-ba3e-1e251a8df962"
/>
Release Notes:
- Added file names and line numbers to symbol completions in the agent
panel
To help make our GitHub Actions easier to understand, we're planning to
split the existing `ci.yml` into three separate workflows:
* run_bundling.yml (this PR)
* run_tests.yml
* make_release.yml
To avoid the duplication that this might otherwise cause, we're planning
to write the workflows with gh-workflow, and use rust instead of
encoding logic in YAML conditions.
Release Notes:
- N/A
---------
Co-authored-by: Ben Kunkle <ben@zed.dev>
Closes#40884
- Make IOPub task return a `Result`
- Create a monitoring task that watches over IOPub, Control, Routing and
Shell tasks.
- If any of these tasks fail, report the error with `kernel_errored()`
(which is already used to report process crashes)
https://github.com/user-attachments/assets/3125f6c7-099a-41ca-b668-fe694ecc68b9
This is not perfect. I did not have time to look into this but:
- When such errors happen, the kernel should be shut down.
- The kernel should no longer appear as online in the UI
But at least the user is getting feedback on what went wrong.
Release Notes:
- Jupyter client errors are now surfaced in the UI (#40884)
After #38882 we were always including file/directory mentions as
`zed:///agent/file?path=a/b/c.rs`.
However, for most resource links (files/directories/symbols/selections)
we want to use a common format, so that ACP servers don't have to
implement custom handling for parsing `ResourceLink`s coming from Zed.
This is what it looks like now:
```
[@index.js](file:///Users/.../projects/reqwest/examples/wasm_github_fetch/index.js)
[@wasm](file:///Users/.../projects/reqwest/src/wasm)
[@Error](file:///Users/.../projects/reqwest/src/async_impl/client.rs?symbol=Error#L2661:2661)
[@error.rs (23:27)](file:///Users/.../projects/reqwest/src/error.rs#L23:27)
```
Release Notes:
- N/A
---------
Co-authored-by: Cole Miller <cole@zed.dev>
Closes https://github.com/zed-industries/zed/issues/41334
This comes down to a caching issue..
Release Notes:
- Fixed an issue where the scrollbar track color would not update in
case the theme was changed.
The current Jetbrains keymap has `ctrl-shift-f12` set to
`CloseAllDocks`. On Jetbrains IDEs this hotkey actually toggles the
docks, which is very convenient: You press it once to hide all docks and
just focus on the code, and then you can press it again to toggle your
docks right back to how they were. Unlike `CloseAllDocks`, a toggle
means the editor needs to remember the previous docks state so this
necessitated some code changes.
Release Notes:
- Added a `Toggle All Docks` editor action and updated the keymaps to
use it
By default, the uninstaller will only delete files that were written by
the original installer. When users upgrade Zed, these new
OpenConsole.exe files will have been written by auto_upgrade_helper, not
the installer. Force them to be deleted on uninstall, so they do not
hang around.
Release Notes:
- N/A
The tab switcher render matches calls each workspace item's
`Item::tab_content` function that can return an element of variable
size. Because the tab switcher was using a uniform list under the hood,
this would cause spacing issues when tab_contents elements had different
sizes.
The fix is by changing the picker to use a material list under the hood.
Release Notes:
- N/A
## Description
Fixes the copy button functionality in REPL interactive mode error
output sections.
When executing Python code that produces errors in the REPL (e.g.,
`NameError`), the copy button in the error output section was
unresponsive. The stdout/stderr copy button worked correctly, but the
error traceback section copy button had no effect when clicked.
Fixes#40207
## Changes
Modified the following:
src/outputs.rs: Fixed context issues in render_output_controls by
replacing cx.listener() with simple closures, and added custom button
implementation for ErrorOutput that copies/opens the complete error
(name + message + traceback)
src/outputs/plain.rs: Made full_text() method public to allow access
from button handlers
src/outputs/user_error.rs: Added Clone derive to ErrorView struct and
removed a couple pieces of commented code
## Why This Matters
The copy button was clearly broken and it is useful to have for REPL
workflows. Users could potentially need to copy error messages for a
variety of reasons.
## Testing
See attached demo for proof that the fix is working as intended. (this
is my first ever commit, if there are additional test cases I need to
write or run, please let me know!)
https://github.com/user-attachments/assets/da158205-4119-47eb-a271-196ef8d196e4
Release Notes:
- Fixed copy button not working for REPL error output
Replace the binary search approach with a more efficient partition_point
method for checking selection overlaps. This eliminates the need to
collect and sort selection ranges separately, reducing memory allocation
and improving performance when handling multiple selections.
Release Notes:
- N/A
Signed-off-by: Xiaobo Liu <cppcoffee@gmail.com>
This change adds two things to our remote server build
process:
1. It now checks if all required tooling is installed before using it or installing it on demand. This includes checks for `rustup` and `cargo-zigbuild` in your `PATH`.
2. Next, if `ZED_BUILD_REMOTE_SERVER` contains `musl` and `ZED_ZSTD_MUSL_LIB`
is set, we will pass its value (the path) to `cargo-zigbuild` as `-C
link-arg=-L{path}`.
Release Notes:
- N/A
The `wrap selections in tag` action currently did not take line_mode
into account, which means when selecting lines with `shift-v`, the
start/end tags would be inserted into the middle of the selection (where
the cursor sits)
https://github.com/user-attachments/assets/a1cbf3da-d52a-42e2-aecf-1a7b6d1dbb32
This PR fixes this behaviour by checking if the selection uses line_mode
and then adjusting start and end points accordingly.
NOTE: I looked into amending the test cases for this, but I am unsure
how to express line mode with range markers. I would appreciate some
guidance on this and then I am happy to add test cases.
After:
https://github.com/user-attachments/assets/a212c41f-b0db-4f50-866f-fced7bc677ca
Release Notes:
- Fixed `Editor: wrap selection in tags` when in vim visual line mode
---------
Co-authored-by: Conrad Irwin <conrad.irwin@gmail.com>
Release Notes:
- Pane key context now includes 'buffer_search_deployed' identifier
The goal of this PR is to add a new identifier in the key context that
will let the user target when the BufferSearchBar is deployed even if
they are not focused on it.
requested in #36930
Same rational as #40454 this will allow users to make more flexible
keybindings, by including some additional information higher up the key
context tree.
i thought adding this context to `Pane` seemed more appropriate than
`Editor` since `Terminal` also has a `BufferSearchBar`; however, I ran
into some import issues between BufferSearchBar, Search, Pane, and
Workspace which made it difficult to implement adding this context
directly inside `Pane`'s render function.
instead i added a new method called `contributes_context` to
`ToolbarItem` which will allow any toolbar item to add additional
context to the `Pane` level, which feels like it might come in handy.
here are some screen shots of the context being displayed in the Editor
and the Terminal
<img width="1653" height="1051" alt="Screenshot 2025-10-25 at 14 34 03"
src="https://github.com/user-attachments/assets/21c5b07a-8d36-4e0b-ad09-378b12d2ea38"
/>
<img width="1444" height="1167" alt="Screenshot 2025-10-25 at 12 32 21"
src="https://github.com/user-attachments/assets/86afe72f-b238-43cd-8230-9cb59fb93b2c"
/>
**Problem:** Current relative line numbering creates a mismatch with
vim-style navigation when soft wrap is enabled. Users must mentally
calculate whether target lines are wrapped segments or logical lines,
making `<n>j/k` navigation unreliable and cognitively demanding.
**How things work today:**
- Real line navigation (`j/k` moves by logical lines): Requires
determining if visible lines are wrapped segments before jumping. Can't
jump to wrapped lines directly.
- Display line navigation (`j/k` moves by display rows): Line numbers
don't correspond to actual row distances for multi-line jumps.
**Proposed solution:** Count and number each display line (including
wrapped segments) for relative numbering. This creates direct
visual-to-navigational correspondence where the relative number shown
always matches the `<n>j/k` distance needed.
**Benefits:**
- Eliminates mental overhead of distinguishing wrapped vs. logical lines
- Makes relative line numbers consistently actionable regardless of wrap
state
- Preserves intuitive "what you see is what you navigate" principle
- Maintains vim workflow efficiency in narrow window scenarios
Also explained an discussed in
https://github.com/zed-industries/zed/discussions/25733.
Release Notes:
Release Notes:
- Added support for counting wrapped lines as relative lines and for
displaying line numbers for wrapped segments. Changes
`relative_line_numbers` from a boolean to an enum: `enabled`,
`disabled`, or `wrapped`.
---------
Co-authored-by: Conrad Irwin <conrad.irwin@gmail.com>
If you set `{"basedpyright": {"analysis": {"typeCheckingMode":
"off"}}}`, you will notice that it doesn't actually work, but
`{"basedpyright.analysis": {"typeCheckingMode": "off"}}` does.
Made the change on how the default is being set.
Release Notes:
- N/A
