Adding a missing conversion for elementID on paint example

[jaredwy]

Relatively new to gpui, but looking at an example on windows and found this compilation error. Hope this is the right way to contribute the patch.

Release Notes:

• Fixed A compilation error in the gpui example painting.

[cla-bot]

We require contributors to sign our Contributor License Agreement, and we don't have @jaredwy on file. You can sign our CLA at https://zed.dev/cla. Once you've signed, post a comment here that says '@cla-bot check'.

[jaredwy]

@cla-bot check

[cla-bot]

The cla-bot has been summoned, and re-checked this pull request!

[SomeoneToIgnore]

Can you elaborate on the "compilation error" a bit?

We do check the examples in CI it seems:

zed/.github/workflows/run_tests.yml

Lines 281 to 288 in af6385a

	- name: cargo build --workspace --bins --examples
	run: cargo build --workspace --bins --examples
	shell: bash -euxo pipefail {0}
	- name: steps::cleanup_cargo_config
	if: always()
	run: |
	rm -rf ./../.cargo
	shell: bash -euxo pipefail {0}

And building the example locally with cargo run -p gpui --example painting works for me.

[jaredwy]

@SomeoneToIgnore I thought it was a bit odd but also couldn't find an implicit conversion anywhere. The error was on the line that's changed, saying no conversion from string to elementIDs. This is also how elementIDs are created in other examples.

The one difference perhaps, is I was not building it inside the workspace, but copied the file into my own project and included gpui that way.

I should have some time this afternoon to try it out inside the project vs out as well as on mac (although can't see why that would be any different)

[SomeoneToIgnore]

Indeed, thank you for the pointer: if I crate a standalone project, do cargo add gpui there and get gpui = "0.2.2", it cannot compile the example's code copied over into main.

I am not sure though, if we should adjust the examples to fix this: sounds like another gpui release is needed instead.

[SomeoneToIgnore]

Oh yeah, we have #46183 and an assignee there.

I believe the real fix would be to publish the newer version?
That would require more than a PR I'm afraid, hence will close the PR.
@\miguelraz should take care of this situation eventually it seems.