codegen-units=1 lets the compiler optimize more aggressively, but it
makes things _much_ slower. The benefits of this has been evaporating
as the compiler has improved, and now that the CLI build is a bottleneck,
let's just remove it.
- Don't set a config option for update interval. We don't let the one
in core be configurable, don't see a huge need for this.
- Don't start the server when the update is available
- Fix some race conditions
- Fix some clippy lints on Rust 1.81
Fixes#191501
It turns out this was a difference in inflate/deflate implementations
between the extension/SDK and the CLI. The SDK uses Node's zlib bindings,
while by default Rust's flate2 library uses a rust port of [miniz][1].
The 'logic' in the CLI was good, but miniz does not appear to flush
decompressed data as nicely on SYNC'd boundaries as zlib does, which
caused data to 'stall'. Telling the flate2 crate to use the native
bindings fixed this.
This could also be the cause of the flakiness occasionally seen on idle
tunnel connections!
[1]: https://github.com/richgel999/miniz
Closes https://github.com/microsoft/vscode/issues/168492
This implements @aeschli's 'server server' concept in a new
`code serve-web` command.
Command line args are similar to the standalone web server. The first
time a user hits that page, the latest version of the VS Code web server
will be downloaded and run. Thanks to Martin's previous PRs, all
resources the page requests are prefixed with `/<quality-<commit>`.
The latest release version is cached, but when the page is loaded again
and there's a new release, a the new server version will be downloaded
and started up.
Behind the scenes the servers all listen on named pipes/sockets and the
CLI acts as a proxy server to those sockets. Servers without connections
for an hour will be shut down automatically.
- Remove the `prepare` script entirely
- Variables are now populated from the product.json during build. Most
variables are mapped automatically, with some special handling in a
few cases. `build.rs` is now much more self-contained.
- Look for the `product.overrides.json` for vscode developers instead of
looking for a peer `vscode-distro` folder
Fixes#178691
* signing: implement signing service on the web
* wip
* cli: implement stdio service
This is used to implement the exec server for WSL. Guarded behind a signed handshake.
* update distro
* rm debug
* address pr comments
* cli: add acquire_cli
As given in my draft document, pipes a CLI of the given platform to the
specified process, for example:
```js
const cmd = await rpc.call('acquire_cli', {
command: 'node',
args: [
'-e',
'process.stdin.pipe(fs.createWriteStream("c:/users/conno/downloads/hello-cli"))',
],
platform: Platform.LinuxX64,
quality: 'insider',
});
```
It genericizes caching so that the CLI is also cached on the host, just
like servers.
* fix bug
Other connected clients will now print:
```
Connected to an existing tunnel process running on this machine. You can press:
- Ctrl+C to detach
- "x" to stop the tunnel and exit
- "r" to restart the tunnel
```
These are then sent to the server to have that take effect. This is
mostly some refactors in the singleton_server to make the lifecycle work.
Fixes#167741
This eschews the offical Windows service system in favor of registering
into `HKEY_CURRENT_USER\Software\Microsoft\Windows\CurrentVersion\Run`.
Unlike services, this can be done without administrative permissions,
does not require the current username/password, and is not blocked by
miscellaneous and mysterious system policies.
Since the process is basically unmanaged by the OS, this requires a
little legwork to start and stop the process when registering and
unregistering.
Adds an stdin/out json rpc server for wsl.
Exposes a singular install_local command to install+boot the vscode server on a port from a local archive.
Also refines the common rpc layer some more. I'm decently happy with it now.
This uses a hash of the tunnel ID to create the connection token, which
should be sufficient to resolve the issues.
We also now publish the protocol version in the tunnel tags, since the
connection token must be supplied in the resolver, which is before we
start connecting to the tunnel.
See https://github.com/microsoft/vscode-internalbacklog/issues/3287
A fully-functioning CLI with tunnel capabilities requires product-specific
configuration that is not included in the OSS. This change has the CLI's build
script automatically look for a peer `vscode-distro` folder and, in a debug
build, set its build variables based on that.
It also works to set correct variables if vscode-distro is not found, based on
the OSS product.json (though this is not sufficient to run a fully fledged
tunnel.)
systemd, like most 'modern' linux components, has a nice dbus API. This uses
that API to register the tunnel as a service of the calling user.
The dbus dependency is temporarily duplicated, until secret-service 3 is
released, where they update to the latest version (should be a week or two).
For https://github.com/microsoft/vscode-cli/issues/367. Next up, macOS,
then it's done :)
The standalone CLI should detect and fall back to using and
system-installed VS Code instance, rather than trying to download zips
and manage its own VS Code instances.
There are three approaches used for discovery:
- On Windows, we can easily and quickly read the register to find
installed versions based on their app ID.
- On macOS, we initially look in `/Applications` and fall back to the
slow `system_profiler` command to list app .app's if that fails.
- On Linux, we just look in the PATH. I believe all Linux installers
(snap, dep, rpm) automatically add VS Code to the user's PATH.
Failing this, the user can also manually specify their installation dir,
using the command `code version use stable --install-dir /path/to/vscode`.
Fixes#164159
