This PR adds a setting to allow configuring the low-speed timeout for
the Assistant when using the OpenAI provider.
The `low_speed_timeout_in_seconds` accepts a number of seconds that the
HTTP client can go below a minimum speed limit (currently set to 100
bytes/second) before it times out.
```json
{
"assistant": {
"version": "1",
"provider": { "name": "openai", "low_speed_timeout_in_seconds": 60 }
},
}
```
This should help the case where the `openai` provider is being used with
a local model that requires higher timeouts.
Issue: https://github.com/zed-industries/zed/issues/9913
Release Notes:
- Added a `low_speed_timeout_in_seconds` setting to the Assistant's
OpenAI provider
([#9913](https://github.com/zed-industries/zed/issues/9913)).
This PR moves the Ruby injections added in #8796 to the right location,
since Ruby support was extracted into an extension in #11360.
Release Notes:
- N/A
This PR extracts Ruby and ERB support into an extension and removes the
built-in Ruby and Ruby support from Zed.
As part of this, the new extension is prepared for adding support for
the `Ruby LSP` which has some blockers. See
https://github.com/zed-industries/zed/pull/8613 I was thinking of adding
an initial support for Ruby LSP but I think it would be better to start
with extracting the Ruby extension for now.
The implementation, as the 1st step, matches the bundled version but
with 3 differences:
1. Added signature output to the completion popup. See my comment below.
3. Use the shell environment for starting the `solargraph` executable.
See my comment below.
4. Bumped the tree sitter version for Ruby to the latest available
version.
Additionally, I plan to tweak this extension a bit in the future but I
think we should do this bit by bit. Thanks!
Release Notes:
- Removed built-in support for Ruby, in favor of making it available as
an extension.
---------
Co-authored-by: Marshall Bowers <elliott.codes@gmail.com>
I ran into this when trying to get #11550 working: the VCS menu would
open repositories on its owned, based on paths, instead of going through
the worktree on which we already store the git repositories.
Release Notes:
- N/A
Release Notes:
- vim: Added support for the changelist. `g;` and `g,` to the
previous/next change
- vim: Added support for the `'.` mark
- vim: Added support for `gi` to resume the previous insert
- [x] Build out cli on linux
- [x] Add support for --dev-server-token sent by the CLI
- [x] Package cli into the .tar.gz
- [x] Link the cli to ~/.local/bin in install.sh
Release Notes:
- linux: Add cli support for managing zed
While these would match how macOS handles this scenario, they crash on
Catalina, and require mouse clicks to interact.
cc @bennetbo
Release Notes:
- N/A
Release Notes:
- vim: Added support for buffer-local marks (`'a-'z`) and some builtin
marks `'<`,`'>`,`'[`,`']`, `'{`, `'}` and `^`. Global marks (`'A-'Z`),
and other builtin marks (`'0-'9`, `'(`, `')`, `''`, `'.`, `'"`) are not
yet implemented. (#5122)
---------
Co-authored-by: Conrad Irwin <conrad.irwin@gmail.com>
To speed up Linux CI builds, use a set of self-hosted Linux machines and
use them to run all slow CI steps for Linux: "tests", bundling and
nightly builds.
Also adds a set of dev icons as Linux bundling script required them for
`run-bundling`-tagged builds from regular PRs.
Same icons as for Preview were used, but, ideally, something different
could be created.
Release Notes:
- N/A
From now on, only top-level captures are treated as runnable tags and
the rest is appended to task context as custom environmental variables
(unless the name is prefixed with _, in which case the capture is
ignored). This is most likely gonna help with Pest-like test runners.
Release Notes:
- N/A
---------
Co-authored-by: Remco <djsmits12@gmail.com>
This PR leverages a more modern Windows API to implement
`WindowsDispatcher`, aligning its implementation more closely with that
of the `macOS` platform. The following improvements have been made:
1. Similar to `macOS`, there is no longer a need to use `sender` and
`receiver` to dispatch a `Runnable` on the main thread.
2. There is also no longer a need to use an `Event` for synchronization.
3. Consistent with #7506 and #11269, `Runnable` is now executed with
high priority.
However, this PR raises the minimum Windows version requirement of
`GPUI` to Windows 10, specifically Windows 10 Fall Creators Update
(10.0.16299). However, the `alacritty_terminal` dependency in Zed relies
on `conPTY` on Windows, an API introduced in the Windows 10 Fall
Creators Update. Therefore, the impact of this PR on Zed should be
minimal. I'd like to hear your voices about this PR, especially about
the minimum Windows version bumping.
Release Notes:
- N/A
There're two errors in scrollbar markers in the presence of folded code:
1. Some markers are not displayed (when the marked row numbers are
greater than the total displayed rows count after folding).
2. Code folding / unfolding doesn't trigger markers repainting.
This PR fixes both problems.
Release Notes:
- Fixed scrollbar markers for folded code.
The second problem (markers are repainted after I move the cursor, not
after folding):
https://github.com/zed-industries/zed/assets/2101250/57ed563d-186d-4497-98ab-d4f946416726
This PR updates the saved conversation picker to use a list of
conversations retrieved from disk instead of the static placeholder
values.
Release Notes:
- N/A
This PR updates the root path used by Copilot to be a validate path when
running on Windows.
Release Notes:
- N/A
Co-authored-by: Jason Lee <huacnlee@gmail.com>
A minor thing I've spotted and decided to fix on the spot.
It was being cloned twice within the body of that function (one of which
was redundant even without this PR); now in most cases we go down from 2
clones to 0.
Release Notes:
- N/A
This pull request introduces a new `markdown` crate which is capable of
parsing and rendering a Markdown source. One of the key additions is
that it enables text selection within a `Markdown` view. Eventually,
this will replace `RichText` but for now the goal is to use it in the
assistant revamped assistant in the spirit of making progress.
<img width="711" alt="image"
src="https://github.com/zed-industries/zed/assets/482957/b56c777b-e57c-42f9-95c1-3ada22f63a69">
Note that this pull request doesn't yet use the new markdown renderer in
`assistant2`. This is because we need to modify the assistant before
slotting in the new renderer and I wanted to merge this independently of
those changes.
Release Notes:
- N/A
---------
Co-authored-by: Nathan Sobo <nathan@zed.dev>
Co-authored-by: Conrad <conrad@zed.dev>
Co-authored-by: Alp <akeles@umd.edu>
Co-authored-by: Zachiah Sawyer <zachiah@proton.me>
It seems that on the first frame after the system resumes from sleep,
`dcomp_vsync_fn` mistakenly detects the `timer_stop_event` triggering
and exits the loop.
Release Notes:
- N/A
Release Notes:
- N/A
## What
Addresses a long-standing issue of doing the blending operations in sRGB
space. Currently, the input HSL colors are provided in sRGB space and
converted to linear in the vertex shader. Conversion back to sRGB, which
is required on most platforms today, happens at the very end of the
pipeline when writing into sRGB render target.
Note-1: in the future we may consider doing HSL -> sRGB -> Linear
transform on CPU before feeding into shaders. However, I don't expect
any significant difference here given that we are likely bound by fill
rate and pixel shaders, anyway.
Note-2: the graphics stack is programmed to detect if the platform
supports presenting in linear color space and avoids converting to sRGB
at the end in this case. However, on my Z13 laptop this isn't supported
by the RADV driver.
Closes#7684Closes#11462
@jansol please confirm if you can!
## Comparison
Screenshot of the Glazier theme before the change:
Same theme after the change:
Previously, `DirectWrite` had been following the text system
implementation on `macOS`, using the font's postscript name to
differentiate between different font faces. However, I noticed
occasional rendering issues, such as fonts incorrectly rendering as
italics. Later, I discovered that on `Windows`, the postscript name is
**not** unique. Surprisingly, even the same font can have different
postscript names on macOS and Windows! It's hard to believe! The
postscript name of a font face should be obtained from the same font
table. Why would the same font face have different names?
For example, the postscript name of ZedMono on `macOS` is
`Zed-Mono-Bold-Extended-Italic`, while on `Windows`, it is
`Zed-Mono-Extended`, missing weight and style information, leading to
incorrect rendering.
This PR introduces a `FontIdentifier` struct to uniquely identify font
faces.
Release Notes:
- N/A
As per Microsoft documentation, positive values scroll right, not left.
GPUI was incorrectly assuming it perfectly mirrored vertical scrolling.
Fixes#11515
Release Notes:
- N/A
This PR is another step to tabless editing (#6424, #4963). It adds
support for tab bar settings that allow the user to change its placement
or to hide completely.
Configuraton:
```json
"tab_bar": {
"show": true
}
```
Placemnet options are "top", "bottom" and "no".
This PR intentionally doesn't affect tab bars of other panes (Terminal
for instance) to keep code changes small. I guess we'll do the rest in
separate PRs.
Release Notes:
- Added support for configuring the editor tab bar (part of #6424,
#4963).
---------
Co-authored-by: Mikayla <mikayla@zed.dev>
This shows "Researching..." as placeholder text as early as possible so
that the user can see the model is working on reading/researching/etc.
This also adds on an `Option<Value>` to the `render_running` function so
that tools can hopefully render based on partially completed JSON (still
to come).
Release Notes:
- N/A
This fixes restart after updates not working on Linux.
On Linux we can't reliably get the binary path after an update, because
the original binary was deleted and the path will contain ` (deleted)`.
See: https://github.com/rust-lang/rust/issues/69343
We *could* strip ` (deleted)` off, but that feels nasty. So instead we
save the original binary path, before we do the installation, then
restart.
Later on, we can also change this to be a _new_ binary path returned by
the installers, which we then have to start.
Release Notes:
- N/A
