* separate saved/runtime structure, kind of working
* serializing config
* work
* work
* save config through the configuration screen
* work
* startup wizard
* style(code): cleanups
* fix(session): reload config from disk when switching sessions
* style(fmt): rustfmt
* fix(config): propagate cli config options to screen
* style(fmt): rustfmt
* add context to opening command panes, and events for their start/finish
* show/hide other panes plugin apis
* get tests to pass
* style(fmt): rustfmt
* update plugin test snapshots
Update `get_cwd()` to only refresh process specifics for the given PID,
matching the `get_cwds()` implementation.
This significantly reduces the latency of opening a new Pane or Tab.
On a 64-core Threadripper system, this reduces the latency of
`get_cwd()` from 850ms to 1ms. On an 8-core Ryzen system, the
improvement is smaller, but still perceptible.
* cargo: Reintroduce `rust-version` field
in `Cargo.toml` since it appears that client-side `cargo` will check
this field and [error when the users rust toolchain is too old][1]. Add
comments in `Cargo.toml` and `rust-toolchain.toml` pointing to each
other so the rust versions mentioned in each file are hopefully kept in
sync in the future.
[1]: https://doc.rust-lang.org/cargo/reference/manifest.html#the-rust-version-field
* CHANGELOG: Add MR #3482.
* cargo: Remove `rust-version` (msrv) field
from `Cargo.toml`. This has gone out of sync with the rust version
specified in `rust-toolchain.toml` and appears to have little
consequence to development, since it defines only the MSRV. We're using
`rust-toolchain.toml` to strictly enforce a specific rust version for
development, hence we stick with that.
Closes#3476
* CHANGELOG: Add MR #3480.
* work
* work
* working
* get default mode from server and some ui responsiveness
* work
* finish design and get tests to pass
* get e2e tests to pass
* add classic layout
* add classic layout assets
* fix e2e tests
* style(fmt): rustfmt
* fix plugin system test
* style(fmt): some cleanups
* Remove ForeignFunctionEnv wrapper around PluginEnv
This will enable PluginEnv to be the Store context when migrating to
Wasmtime.
* Pass PluginEnv by value to load_plugin_instance
This will allow removing the Clone impl from PluginEnv when migrating to
Wasmtime as required by the missing Clone impl on Wasmtime's WasiCtx.
* Avoid passing a Store around when an Engine is enough
* Pass PluginEnv to the wasi read/write functions
Wasmtime requires storing the read/write end of the pipe outside of the
WasiCtx. Passing PluginEnv to these functions allows storing them in the
PluginEnv.
* Migrate to Wasmtime
* Switch from wasi-common to wasmtime-wasi
* Reduce verbosity of wasmtime_wasi logs
* Increase startup delay
To wait for all plugins to be compiled.
* Disable some wasmtime features
* Update to Wasmtime 21.0.1
* refactor(server): interpret keys on server so they can be rebound
* feat(plugins): allow rebinding keys at runtime
* various cleanups
* add tests
* style(fmt): rustfmt
* fix(tests): address (some) e2e test flakiness
* style(fmt): rustfmt
