add try_insert to entity commands by ethereumdegen · Pull Request #9844 · bevyengine/bevy

ethereumdegen · 2023-09-19T00:22:33Z

Objective

I spoke with some users in the ECS channel of bevy discord today and they suggested that I implement a fallible form of .insert for components.
In my opinion, it would be nice to have a fallible .insert like .try_insert (or to just make insert be fallible!) because it was causing a lot of panics in my game. In my game, I am spawning terrain chunks and despawning them in the Update loop. However, this was causing bevy_xpbd to panic because it was trying to .insert some physics components on my chunks and a race condition meant that its check to see if the entity exists would pass but then the next execution step it would not exist and would do an .insert and then panic. This means that there is no way to avoid a panic with conditionals.

Luckily, bevy_xpbd does not care about inserting these components if the entity is being deleted and so if there were a .try_insert, like this PR provides it could use that instead in order to NOT panic.

( My interim solution for my own game has been to run the entity despawn events in the Last schedule but really this is just a hack and I should not be expected to manage the scheduling of despawns like this - it should just be easy and simple. IF it just so happened that bevy_xpbd ran .inserts in the Last schedule also, this would be an untenable soln overall )

Solution

Describe the solution used to achieve the objective above.

Add a new command named TryInsert (entitycommands.try_insert) which functions exactly like .insert except if the entity does not exist it will not panic. Instead, it will log to info. This way, crates that are attaching components in ways which they do not mind that the entity no longer exists can just use try_insert instead of insert.

Changelog

Additional Thoughts

In my opinion, NOT panicing should really be the default and having an .insert that does panic should be the odd edgecase but removing the panic! from .insert seems a bit above my paygrade -- although i would love to see it. My other thought is it would be good for .insert to return an Option AND not panic but it seems it uses an event bus right now so that seems to be impossible w the current architecture.

github-actions · 2023-09-19T00:22:47Z

Welcome, new contributor!

Please make sure you've read our contributing guide and we look forward to reviewing your pull request shortly ✨

nicopap

Looks good, I think we should merge this. I'd like to see minor adjustments though.

nicopap · 2023-09-19T08:20:20Z

crates/bevy_ecs/src/system/commands/mod.rs

+    ///
+    /// fn add_combat_stats_system(mut commands: Commands, player: Res<PlayerEntity>) {
+    ///     
+    ///         commands.entity(player.entity)


Suggested change

/// commands.entity(player.entity)

/// commands.entity(player.entity)

nicopap · 2023-09-19T08:20:35Z

crates/bevy_ecs/src/system/commands/mod.rs

+    ///             health: Health(100),
+    ///             strength: Strength(40),
+    ///         });
+    ///     


For the lulz (and illustration) I'd add this line

Suggested change

///

/// commands.entity(player.entity).despawn();

Edit: Actually this should be moved before the insert.

nicopap · 2023-09-19T08:24:06Z

crates/bevy_ecs/src/system/commands/mod.rs

+        if let Some(mut entity) = world.get_entity_mut(self.entity) {
+            entity.insert(self.bundle);
+        } else {
+            info!("error[B0003]: Could not insert a bundle (of type `{}`) for entity {:?} because it doesn't exist in this World.", std::any::type_name::<T>(), self.entity);


Suggested change

info!("error[B0003]: Could not insert a bundle (of type `{}`) for entity {:?} because it doesn't exist in this World.", std::any::type_name::<T>(), self.entity);

debug!("[B0003]: Could not insert a bundle (of type `{}`) for entity {:?} because it doesn't exist in this World.", std::any::type_name::<T>(), self.entity);

joseph-gio · 2023-09-20T04:07:28Z

crates/bevy_ecs/src/system/commands/mod.rs

+        if let Some(mut entity) = world.get_entity_mut(self.entity) {
+            entity.insert(self.bundle);
+        } else {
+            info!("error[B0003]: Could not insert a bundle (of type `{}`) for entity {:?} because it doesn't exist in this World.", std::any::type_name::<T>(), self.entity);


This log seems like unnecessary noise. If someone calls try_insert that means they don't care about this "error" case, so we can safely ignore it.

joseph-gio · 2023-09-20T04:12:49Z

crates/bevy_ecs/src/system/commands/mod.rs

+    ///
+    /// # Note
+    ///
+    /// The command will not panic if the associated entity does not exist.


Suggested change

/// The command will not panic if the associated entity does not exist.

/// Unlike [`Self::insert`], this command will not panic if the associated entity does not exist.

ethereumdegen · 2023-09-20T13:44:32Z

okay updated based on fdbk

nicopap

LGTM

crates/bevy_ecs/src/system/commands/mod.rs

# Objective - I spoke with some users in the ECS channel of bevy discord today and they suggested that I implement a fallible form of .insert for components. - In my opinion, it would be nice to have a fallible .insert like .try_insert (or to just make insert be fallible!) because it was causing a lot of panics in my game. In my game, I am spawning terrain chunks and despawning them in the Update loop. However, this was causing bevy_xpbd to panic because it was trying to .insert some physics components on my chunks and a race condition meant that its check to see if the entity exists would pass but then the next execution step it would not exist and would do an .insert and then panic. This means that there is no way to avoid a panic with conditionals. Luckily, bevy_xpbd does not care about inserting these components if the entity is being deleted and so if there were a .try_insert, like this PR provides it could use that instead in order to NOT panic. ( My interim solution for my own game has been to run the entity despawn events in the Last schedule but really this is just a hack and I should not be expected to manage the scheduling of despawns like this - it should just be easy and simple. IF it just so happened that bevy_xpbd ran .inserts in the Last schedule also, this would be an untenable soln overall ) ## Solution - Describe the solution used to achieve the objective above. Add a new command named TryInsert (entitycommands.try_insert) which functions exactly like .insert except if the entity does not exist it will not panic. Instead, it will log to info. This way, crates that are attaching components in ways which they do not mind that the entity no longer exists can just use try_insert instead of insert. --- ## Changelog ## Additional Thoughts In my opinion, NOT panicing should really be the default and having an .insert that does panic should be the odd edgecase but removing the panic! from .insert seems a bit above my paygrade -- although i would love to see it. My other thought is it would be good for .insert to return an Option AND not panic but it seems it uses an event bus right now so that seems to be impossible w the current architecture.

add try insert

9c73ad1

ethereumdegen changed the title ~~add try insert~~ add try_insert to entity commands Sep 19, 2023

alice-i-cecile added A-ECS Entities, components, systems, and events C-Usability A targeted quality-of-life change that makes Bevy easier to use labels Sep 19, 2023

ethereumdegen added 2 commits September 18, 2023 22:23

fix notes

fcaf1fc

fmt

053ed5a

nicopap reviewed Sep 19, 2023

View reviewed changes

joseph-gio reviewed Sep 20, 2023

View reviewed changes

fixing comments

10c30e6

nicopap approved these changes Sep 20, 2023

View reviewed changes

crates/bevy_ecs/src/system/commands/mod.rs Show resolved Hide resolved

ethereumdegen added 2 commits September 20, 2023 09:53

fixing docs

914090a

fmt

7f56c1b

joseph-gio approved these changes Sep 20, 2023

View reviewed changes

joseph-gio added the S-Ready-For-Final-Review This PR has been approved by the community. It's ready for a maintainer to consider merging it label Sep 20, 2023

james7132 approved these changes Sep 20, 2023

View reviewed changes

james7132 added this pull request to the merge queue Sep 20, 2023

Merged via the queue into bevyengine:main with commit 3ee9edf Sep 20, 2023

alice-i-cecile mentioned this pull request Oct 10, 2023

Don't panic when applying an Insert command #9379

Closed

ChristopherBiscardi mentioned this pull request Dec 30, 2023

EntityCommands should have methods to fail gracefully #3845

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

add try_insert to entity commands#9844

add try_insert to entity commands#9844
james7132 merged 6 commits intobevyengine:mainfrom
ethereumdegen:try-insert

ethereumdegen commented Sep 19, 2023 •

edited

Loading

Uh oh!

github-actions bot commented Sep 19, 2023

Uh oh!

nicopap left a comment

Uh oh!

nicopap Sep 19, 2023

Uh oh!

nicopap Sep 19, 2023 •

edited

Loading

Uh oh!

nicopap Sep 19, 2023

Uh oh!

joseph-gio Sep 20, 2023

Uh oh!

joseph-gio Sep 20, 2023

Uh oh!

ethereumdegen commented Sep 20, 2023

Uh oh!

nicopap left a comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

	/// commands.entity(player.entity)
	/// commands.entity(player.entity)

	info!("error[B0003]: Could not insert a bundle (of type `{}`) for entity {:?} because it doesn't exist in this World.", std::any::type_name::<T>(), self.entity);
	debug!("[B0003]: Could not insert a bundle (of type `{}`) for entity {:?} because it doesn't exist in this World.", std::any::type_name::<T>(), self.entity);

	/// The command will not panic if the associated entity does not exist.
	/// Unlike [`Self::insert`], this command will not panic if the associated entity does not exist.

Uh oh!

Conversation

ethereumdegen commented Sep 19, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Objective

Solution

Changelog

Additional Thoughts

Uh oh!

github-actions bot commented Sep 19, 2023

Uh oh!

nicopap left a comment

Choose a reason for hiding this comment

Uh oh!

nicopap Sep 19, 2023

Choose a reason for hiding this comment

Uh oh!

nicopap Sep 19, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

nicopap Sep 19, 2023

Choose a reason for hiding this comment

Uh oh!

joseph-gio Sep 20, 2023

Choose a reason for hiding this comment

Uh oh!

joseph-gio Sep 20, 2023

Choose a reason for hiding this comment

Uh oh!

ethereumdegen commented Sep 20, 2023

Uh oh!

nicopap left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

ethereumdegen commented Sep 19, 2023 •

edited

Loading

nicopap Sep 19, 2023 •

edited

Loading