Config Reverting `[SYNTH-201]` by rutmanz · Pull Request #1425 · Autodesk/synthesis

rutmanz · 2026-07-17T19:22:38Z

Task

Symptom

Previously, config reverting was somewhat suspicious and didn't always revert everything as would be expected

Solution

Added a registerCleanupCallback function to all configuration subpanels, which allows revert callbacks to be applied locally within each subpanel. This means each subpanel is responsible for reverting its own changes, making it much clearer what's going on. The system also supports confirm callbacks, but as most changes are applied immediately, this is only used for some memory freeing right now.

Verification

Spawn an asset, change config things, and try to find a way to have changes not revert on cancel.

Before merging, ensure the following criteria are met:

All acceptance criteria outlined in the ticket are met.
Necessary test cases have been added and updated.
A feature toggle or safe disable path has been added (if applicable).
User-facing polish:
- Ask: "Is this ready-looking?"
Cross-linking between Jira and GitHub:
- PR links to the relevant Jira issue.
- Jira ticket has a comment referencing this PR.

0xda157

there's a lot going on here that seems unrelated to the goal of the pr and makes it quite difficult to review imo. splitting this up into a few commits would make me a lot more confident about reviewing it

0xda157 · 2026-07-17T20:56:49Z

+import type SynthesisBrain from "@/systems/simulation/synthesis_brain/SynthesisBrain.ts"
+import type WPILibBrain from "@/systems/simulation/wpilib_brain/WPILibBrain.ts"

 export type BrainType = "synthesis" | "wpilib" | "unknown"


is braintype ever unknown? I only seen WPILibBrain and SynthesisBrain as implementations of this class.

rutmanz · 2026-07-17T21:17:12Z

Not sure what you want me to do about that. I agree it would make more sense as multiple commits, but just due to how I was working on this that didn't make sense to me in the moment (I kept getting interrupted and doing things out of order). It's something I'll be more mindful of going forward, but I also don't think it makes a lot of sense to try and retroactively split into multiple commits because the intent is already lost.

I had Claude split my commit up into features (as it saw fit). I checked that the diff is the same, and it should be easier for you to review. The groups were made by Claude though, so they are a little off, but I don't think putting more effort into that is a good use of time.

rutmanz · 2026-07-17T21:29:35Z

oh didn't mean to close this let me work on that

Rename the public field to a private `_brainIndexSchemeMap` with a public getter, and add `getBrainIndexSchemeMapping()` alongside the existing `setBrainIndexSchemeMapping()` so all map access goes through explicit methods.

Replace `instanceof` and `brainType ===` checks with narrowing type guards on Brain. Also rename `getPositionTransform` to `getXZPositionTransform` for clarity, add `getRootBody()` helper to MirabufSceneObject, and remove the AutoTestPanel shortcut from the WPILib context menu.

Makes SelectMenu<OptionType> type-safe so callers no longer need to cast `val as SomeSelectionOption`. Removes redundant key prop from ItemCard in ImportMirabufPanel and updates WiringPanel accept text.

Add CleanupRegisterFunction, ConfigurationSubpanelProps, and ConfigurationSubpanelComponent types that all config subpanels will implement. Rename SUBSYSTEMS→JOINT_SUBSYSTEMS and SEQUENTIAL→JOINT_SEQUENCE for clarity. Move the per-configuration-type mode lists (synthesisRobotConfigModes, wpilibRobotConfigModes, fieldConfigModes) out of ConfigurePanel into ConfigTypes.

… system Remove the monolithic ConfigInterface switch and the per-panel originalPrefs refs. Replace with a registerCleanupFunction callback that each subpanel calls to register its own apply/revert logic. ConfigurePanel accumulates these and runs them in FIFO order on accept or LIFO on cancel. Extract MoveInterface and ControlsConfigInterface as standalone subpanel components. Add ConfirmChangesModal for the case where the user navigates away from Controls with unsaved changes.

…a panels Convert each panel to ConfigurationSubpanelComponent and register a revert callback that restores the original value when the user cancels.

Convert ConfigureGamepiecePickupInterface and ConfigureShotTrajectoryInterface to ConfigurationSubpanelComponent and register revert callbacks that restore intake/ejector preferences on cancel.

Convert ConfigureSubsystemsInterface and SequentialBehaviorsInterface to ConfigurationSubpanelComponent with revert callbacks. Fix SubsystemRowInterface to track invertMotor state locally and restrict the unstick-force slider to the drivetrain group only.

Convert ConfigureScoringZonesInterface, ConfigureProtectedZonesInterface, and ConfigureCameraPointsInterface to ConfigurationSubpanelComponent. Each registers a revert callback that restores the original zone/camera point lists and re-syncs the scene on cancel.

Convert SimulationInterface to ConfigurationSubpanelComponent and register a revert callback for SimAutoReconnect. Thread registerCleanupFunction through ConfigureInputsInterface and ConfigureSchemeInterface so input binding changes can also be reverted on cancel.

…ting-split-up

AlexD717

The inputs configuration do not revert when opening a panel that closes the configure panel. Also if a panel were to close the configure panel the ConfirmChangesModal should probably open.

Screen.Recording.2026-07-22.at.1.20.37.PM.mov

rutmanz · 2026-07-23T17:02:47Z

These can't use the "correct" react pattern of setting to new objects because I'm trying to mutate the underlying objects that selectedAssembly and selectedScheme reference. Ideally they would be keyed and the react state is based on that key, but I think that's out of scope for this PR (SYNTH-251 tracks this for schemes)

rutmanz · 2026-07-24T18:09:18Z

@AlexD717, rather than try to interrupt the panel closing process, which would be rather obnoxious, I reframed the language and behavior so that save is the default and reverting is explicitly "revert" instead of "cancel." Does that seem like an appropriate way to deal with this?

AlexD717 · 2026-07-24T18:17:01Z

That seems reasonable

AlexD717

lotm

Note: there are some merge conflicts

rutmanz requested review from a team as code owners July 17, 2026 19:22

rutmanz requested review from 0xda157, AlexD717, PepperLola, azaleacolburn and elliefre and removed request for elliefre July 17, 2026 19:22

0xda157 requested changes Jul 17, 2026

View reviewed changes

rutmanz closed this Jul 17, 2026

rutmanz deleted the zachr/201/config-reveting branch July 17, 2026 21:23

rutmanz restored the zachr/201/config-reveting branch July 17, 2026 21:38

rutmanz reopened this Jul 17, 2026

rutmanz added 11 commits July 17, 2026 17:49

refactor(InputSystem): encapsulate brainIndexSchemeMap access

fb784cd

Rename the public field to a private `_brainIndexSchemeMap` with a public getter, and add `getBrainIndexSchemeMapping()` alongside the existing `setBrainIndexSchemeMapping()` so all map access goes through explicit methods.

refactor(SelectMenu): add generic type parameter

45e7988

Makes SelectMenu<OptionType> type-safe so callers no longer need to cast `val as SomeSelectionOption`. Removes redundant key prop from ItemCard in ImportMirabufPanel and updates WiringPanel accept text.

feat(config): add revert support to alliance/brain/drivetrain/metadat…

040e2b4

…a panels Convert each panel to ConfigurationSubpanelComponent and register a revert callback that restores the original value when the user cancels.

feat(config): add revert support to intake and ejector panels

1469be1

Convert ConfigureGamepiecePickupInterface and ConfigureShotTrajectoryInterface to ConfigurationSubpanelComponent and register revert callbacks that restore intake/ejector preferences on cancel.

Merge remote-tracking branch 'origin/dev' into zachr/201/config-rever…

68ac376

…ting-split-up

rutmanz force-pushed the zachr/201/config-reveting branch from 4bf06af to 68ac376 Compare July 18, 2026 01:40

rutmanz requested a review from a team as a code owner July 18, 2026 01:40

rutmanz requested a review from 0xda157 July 20, 2026 15:17

chore: remove debug prints and style for iterators

cb29068

Merge remote-tracking branch 'origin/dev' into zachr/201/config-reveting

e7cc437

AlexD717 assigned rutmanz Jul 21, 2026

Dhruv-0-Arora reviewed Jul 21, 2026

View reviewed changes

Comment thread ...n/src/ui/panels/configuring/assembly-config/interfaces/ConfigureGamepiecePickupInterface.tsx Outdated

Comment thread fission/src/mirabuf/MirabufSceneObject.ts

Dhruv-0-Arora requested changes Jul 21, 2026

View reviewed changes

Comment thread fission/src/ui/panels/configuring/assembly-config/ConfigurePanel.tsx Outdated

fix: add back reverting to ejector and rename intake and ejector files

f472c32

rutmanz requested a review from Dhruv-0-Arora July 21, 2026 17:42

chore: run biome

5e5601f

Dhruv-0-Arora approved these changes Jul 21, 2026

View reviewed changes

AlexD717 requested changes Jul 22, 2026

View reviewed changes

0xda157 reviewed Jul 23, 2026

View reviewed changes

rutmanz requested a review from 0xda157 July 23, 2026 17:01

rutmanz added 3 commits July 24, 2026 09:32

Merge remote-tracking branch 'origin/dev' into zachr/201/config-reveting

89c9c4d

fix: make mutually exclusive panels fire close events for each other

a9ddf1b

feat: make "save" the default and show toast on panel override

47012a4

rutmanz requested a review from AlexD717 July 24, 2026 18:08

feat: show toast on reverted config

28a2c4d

rutmanz force-pushed the zachr/201/config-reveting branch from 2e78e6e to 28a2c4d Compare July 24, 2026 18:10

AlexD717 approved these changes Jul 24, 2026

View reviewed changes

Merge remote-tracking branch 'origin/dev' into zachr/201/config-reveting

eebd08c

PepperLola added ui/ux Relating to user interface, or in general, user experience gameplay Relating to the playability of Synthesis refactor The most important part of software development. labels Jul 24, 2026

Uh oh!

Conversation

rutmanz commented Jul 17, 2026

Task

Symptom

Solution

Verification

Uh oh!

0xda157 left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

0xda157 Jul 17, 2026

Choose a reason for hiding this comment

Uh oh!

rutmanz commented Jul 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

rutmanz commented Jul 17, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

AlexD717 left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

rutmanz commented Jul 23, 2026

Uh oh!

rutmanz commented Jul 24, 2026

Uh oh!

AlexD717 commented Jul 24, 2026

Uh oh!

AlexD717 left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

rutmanz commented Jul 17, 2026 •

edited

Loading