* Fix vite build (tauri build still broken)
* Fix yarn builds with a couple of shortcuts
* Fix file creation
* Fix documentDir behavior
* Got stream with default file
* Clean up
* Clean up
* Use 'unstable'; fix devtools callsite
The API changed a bit here, which forces us to use the unstable crate
feature. The call to open devtools is also new; it's now on the
webview not window.
Signed-off-by: Paul R. Tagliamonte <paul@kittycad.io>
* Bring back read_dir_recursive from v1
* Fix dates
* More fixes, incl. conf files
* cargo fmt
* Add Updater plugin
* Fix types
* Fix isTauri detection and updater bootup
* Schemas
* Clean up
* Disable devtools
* Attempt at fixing builds
* WIP Ubuntu dep
* WIP Ubuntu dep
* WIP keys in debug
* Enable updater only on release builds
* Reenable webtools on debug
* No linux bundles
* Typo
* Attemp at fixing --bundles none
* Manual tauri debug build
* Empty commit to trigger the CI
* Fix settings
* Empty commit to trigger the CI
* Merge branch 'main' into pierremtb/issue1349
* Add allow-create perm
* tauri-driver no cap
* Empty commit to trigger the CI
* Clean up
* Clean up
* Migrate to tauri v2
Fixes#1349
* Fix fmt
* Merge branch 'main' into pierremtb/issue1349
* Force BUILD_RELEASE: true
* Bump tauri to new beta
* Merge branch 'main' into pierremtb/issue1349
* Fix linux tests
* Fix types
* Add --verbose to tauri-action
* Move --verbose to front
* Back to tauri-driver@0.1.3 and single \ for win
* Back to latest driver, and windows wip
* Disable release conf temporarily
* Rollback to 2.0.0-beta.2
* Rollback to 2.0.0-beta.1
* Move bundle to root for src-tauri/tauri.release.conf.json
* All packages to latest (add http and shell to package.json)
* Testing latest commit for tauri-action
* Remove tauri action
* Add cat
* WIP
* Update ci.yml
* Disable release conf
* Disable rust cache
* Add tauri-action back for release builds
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* Update .codespellrc
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* Trigger CI
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* Fix type
* Clean up
* More clean up
* Fix path concatenation with join
* Attempt at fixing linux tests
* Config clean up
* Downgrade to tauri-driver@0.1.3
* Looks like tauri v2 is actually doing better with linux package names ah!
* Change Linux apt packages
* Increase wdio connectionRetryTimeout
* Revert connectionRetryTimeout and bump tauri packages
* Back to latest tauri-driver
* Disable linux e2e tests
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* Trigger CI
* Clean up
* Update snapshots
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* Trigger CI
* Remove @sentry/react
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* Rename migrated.json to desktop.json
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* Trigger CI
* Change wasm url to http on Windows
---------
Signed-off-by: Paul R. Tagliamonte <paul@kittycad.io>
Co-authored-by: Paul R. Tagliamonte <paul@kittycad.io>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: Adam Chalmers <adam.chalmers@zoo.dev>
* Clean up batch code
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* Remove 'flush_batch: bool' from send_modeling_cmd
It was always being set with false, and it was
bugged for true. If true was set, the cmd would
never actually be run.
* Fix derive-docs
---------
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Send a WsMsg::Close before we bail normally.
This will hopefully trigger engine-manager to release the engine back to
the pool faster, allowing us to increase the number of threads we can
run the modeling-app tests with.
Signed-off-by: Paul R. Tagliamonte <paul@zoo.dev>
* Fix project directory setting input
* Remove unused imports
* Almost working Tauri test
* Finish Tauri e2e test
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* fmt
* Try a different Webriver selector
* Update themeColor component to use new updateValue API
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* Rerun CI
---------
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
* Add a setting for themeColor
* Add primary-color to Tailwind, driven by themeColor setting
* Get rid of most uses of "energy" colors
* Change out the rest of the energy colors
* Tweak NetworkHealthIndicator light mode checkmarks
* Handful of other CSS tweaks while I'm here:
- remove the AppHeader bg and border
- pane margins
- better dark mode button styles
* Make Zoo logomark a badge
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* Re-run CI post-snapshots
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* Retrigger CI
---------
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
* Make it possible to include a setting only on the Settings dialog, not also in the command bar.
* Add web-only setting to permanently dismiss banner
* Honor the dismiss web banner setting
* Remove unused state from useStore
* Make the banner only appear in production builds again
* Print WebSocket errors when we get them
Previously, we eat them and ignore them, but now we'll seek out
and actively print error messages to stderr. We'd previously get a
websocket closed error, but we'll usually get an Error message over the
WebSocket before its closed on us.
Here's some example output during a crash
```
got ws error: WebSocket protocol error: Connection reset without closing handshake
Caused by:
Connection reset without closing handshake
thread 'serial_test_cube_mm' panicked at tests/executor/main.rs:1136:10:
called `Result::unwrap()` on an `Err` value: engine: KclErrorDetails { source_ranges: [SourceRange([180, 188])], message: "Modeling command failed: websocket closed early" }
test serial_test_cube_mm ... FAILED
got error message: {
"error_code": "bad_request",
"message": "Too many active connections, only 2 allowed per user."
}
```
* Rename cargo-criterion to cargo-bench
* Use iai not criterion in CI
We want to benchmark the KCL parser and tokenizer to make sure we don't
accidentally slow them down. Generally Rust projects use Criterion to
benchmark code. Criterion runs your functions a few thousand times to
get reliable wall-clock measurements.
This is good for locally benchmarking but bad for benchmarking in CI.
Why? Because in CI, you're running a container on some shared VM, so
wall-clock time might have a lot of interference from noisy neighbours.
Also, your benchmarks take a long time to run and eat up paid CI minutes.
A better approach for benchmarking in CI is to just count the number of
CPU instructions executed. This correlates with wall-clock time, but it
only needs to run the function once, so it takes much less time. It also
isn't changed by any noisy neighbours running on the same VM or hardware.
This PR adds a new benchmark suite which counts instructions using `iai`,
from the creator of criterion. He says iai and criterion complement each
other nicely. We can run criterion locally and run iai in CI.
* Update image in markdown docs
* dynamic cursor depending on mouse scene state
* hover stuff
* bump min length
* clean up
* sketch on face failing randomly
* more time out for extrude snapshots
* Update src/clientSideScene/sceneEntities.ts
Co-authored-by: Frank Noirot <frank@zoo.dev>
* move for profileStart handle, and select when no tool equiped
---------
Co-authored-by: Frank Noirot <frank@zoo.dev>
* get branch up to where it was before
* setup dots properly
* only show extra handle on hover
* use partical texture for plus button
* fix regression
* fix deleted line
* fix sketch on face test
* caluclate segment length in screen-space/in-pixels
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* side small segment handles on resize
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* Make sure this works on setup and update of segments
* Add to tangential arcs
* Revert "A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)"
This reverts commit 5dc1adacae.
* Revert "A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)"
This reverts commit b8ceea179c.
* try and fix sketch on face in CI
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* more test fix
* convert scaling to be based on pixels
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* trigger ci
* Revert "A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)"
This reverts commit 6287c943dd.
* Revert "A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)"
This reverts commit 1baa3819db.
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* Update src/clientSideScene/segments.ts
Co-authored-by: Frank Noirot <frank@zoo.dev>
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* reduce line thickness
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* trigger CI
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* trigger CI
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* trigger CI
* try putting init script back in
---------
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: Frank Noirot <frank@zoo.dev>
* fix member expression in object expression
Signed-off-by: Jess Frazelle <github@jessfraz.com>
* new renders
Signed-off-by: Jess Frazelle <github@jessfraz.com>
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* empty trigger
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* empty
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
---------
Signed-off-by: Jess Frazelle <github@jessfraz.com>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
* BROKEN: start of scopes for each setting
* Clean up later: mostly-functional scoped settings!
Broken command bar, unimplemented generated settings components
* Working persisted project settings in-folder
* Start working toward automatic commands and settings UI
* Relatively stable, settings-menu-editable
* Settings persistence tweaks after merge
* Custom settings UI working properly, cleaner types
* Allow boolean command types, create Settings UI for them
* Add support for option and string Settings input types
* Proof of concept settings from command bar
* Add all settings to command bar
* Allow settings to be hidden on a level
* Better command titles for settings
* Hide the settings the settings from the commands bar
* Derive command defaultValue from *current* settingsMachine context
* Fix generated settings UI for 'options' type settings
* Pretty settings modal 💅
* Allow for rollback to parent level setting
* fmt
* Fix tsc errors not related to loading from localStorage
* Better setting descriptions, better buttons
* Make displayName searchable in command bar
* Consolidate constants, get working in browser
* Start fixing tests, better types for saved settings payloads
* Fix playwright tests
* Add a test for the settings modal
* Add AtLeast to codespell ignore list
* Goofed merge of codespellrc
* Try fixing linux E2E tests
* Make codespellrc word lowercase
* fmt
* Fix data-testid in Tauri test
* Don't set text settings if nothing changed
* Turn off unimplemented settings
* Allow for multiple "execution-done" messages to have appeared in snapshot tests
* Try fixing up snapshot tests
* Switch from .json to .toml settings file format
* Use a different method for overriding the default units
* Try to force using the new common storage state in snapshot tests
* Update tests to use TOML
* fmt and remove console logs
* Restore units to export
* tsc errors, make snapshot tests use TOML
* Ensure that snapshot tests use the basicStorageState
* Re-organize use of test.use()
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* Update snapshots one more time since lighting changed
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* Fix broken "Show in folder" for project-level settings
* Fire all relevant actions after settings reset
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* Properly reset the default directory
* Hide settings by platform
* Actually honor showDebugPanel
* Unify settings hiding logic
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* fix first extrusion snapshot
* another attempt to fix extrustion snapshot
* Rerun test suite
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* trigger CI
* more extrusion stuff
* Replace resetSettings console log with comment
---------
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: Kurt Hutten Irev-Dev <k.hutten@protonmail.ch>
* set selection as top level event only
* target quirk fix
* fix old bug
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* trigger ci
---------
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Bug: You can see here that the two programs under tests/ are equivalent, just one uses
pipelines and one always assigns to a new sketchgroup. However, the pipeline
produces weird visual bugs. Jess did a git bisect to figure out this was the problem that
Mike was experiencing, around weird visual artifacts with filleting.
Ultimately the bug was that my rewritten `execute_pipe_body` function was executing
the first expression of the pipeline body twice! In most unit tests this didn't matter,
because the first expression in a pipeline was startSketchAt. No big deal to run that
twice. However, in Mike's program, the first expression was `make_circle` or `pentagon`,
user-defined functions that sent a lot of API calls. This meant the pipeline duplicated a lot
of geometry, causing Z-fighting and weird artifacts.
On `main`, the new test program in `tests/executor` causes a stack overflow. Running a flamegraph via `sudo cargo flamegraph --test executor -- serial_test_mike_stress_lines` shows that the problem: `fn execute_pipe_body` is very recursive.
When there's a long pipe, `execute_pipe_body` executes the next child expression, which is always a CallExpression. So `execute_pipe_body` calls `CallExpression::execute`. Which then calls `execute_pipe_body` again.
Fix is simple: `execute_pipe_body` now iterates over pipe subexpressions instead of recursing. This fixes the stack overflow, and is much faster too.
Closes https://github.com/KittyCAD/modeling-app/issues/1891.
