Update onboarding bracket to be robust

[jgomez720]

Updated the onboarding bracket to be more robust. Now it:

• has better defined parameters so when users change values that bracket doesn't fail or look unrealistic
• has asserts to check lengths/widths to make sure output is possible
• Updated sketch/extrude vars to be more descriptive for TTC
• has @settings for units

Just want some ME feedback if this would be understood to others. assert is going to confuse designers, but this is a lesson I'd rather them learn

[qa-wolf]

QA Wolf here! As you write new code it's important that your test coverage is keeping up.
Click here to request test coverage for this PR!

[vercel]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name	Status	Preview	Updated (UTC)
modeling-app	✅ Ready (Inspect)	Visit Preview	Apr 11, 2025 6:43pm

[github-actions]

src/lib/exampleKcl.ts has been updated in this PR, please review and update the src/routes/onboarding, if needed.

[nicboone8]

Is using the bend radius as the fillet radius confusing, either to a user or to T2C?

We'd talked about breaking the onboarding up into smaller tutorials, each focused on a different sample. That would give us some room to start users off with something easy before throwing a ton of new concepts at them. I think using assert is totally fine if it ends up in tutorial 3 or 4-ish

[jgomez720]

Yeah that could be confusing. I'll make it more clear.

For the onboarding, I agree with you in giving users information piece by piece.
cc @franknoirot

[franknoirot]

This closes #4827

[franknoirot]

Yup @nicboone8 @jgomez720 I have #5986 on my task list to support multiple onboarding flows. I think the code itself could include assert but save explaining it for a later more advanced tutorial

[nickmccleery]

This is dope, I didn't even know we had assert statements in KCL. That's a really nice feature.

[pierremtb]

Nice! Will have to make a follow-up PR to #5946 to ensure we can edit with the radius of non-pipe fillets like we have here with the right-click > Edit flow. Just created the issue at #6119

[jessfraz]

I think this broke a test, also maybe one of the front end engineers could help you , we should load this from the bracket/main.kcl versus having two files, thats annoying cc @franknoirot @pierremtb who might be able to help

[jessfraz]

the export test failure seems real so while someone is fixing that we might as well fix how bracket is loaded

[jessfraz]

I can help you fix this tomorrow josh

[pierremtb]

Hit the update button cause the export flakiness was improved with 45e5b25 last night

[jessfraz]

ah okay i still wanna fix the bracket being two places but we can do out of band

[jessfraz]

my ocd ccant handle it

[pierremtb]

Updating as #6164 went in and should fix some if not all issues here

[pierremtb]

Crap I have a feeling e2e/playwright/fixtures/bracket.ts fails on windows and that leads to those 3 real failures

[pierremtb]

I'm on windows now so I can help debug this

[jessfraz]

@pierremtb yeah it looked like it was reading the file but something about the noNewLinesBracket is different

[pierremtb]

I think I was wrong, the tests that were failing at https://github.com/KittyCAD/modeling-app/actions/runs/14281282402/job/40032997357 are all passing locally on my windows laptop 🤔

[jessfraz]

oh thats good maybe ci is just being a butt

[jgomez720]

dear god what happened to this

[pierremtb]

Yeah those tests are busted in CI, will have another look tomorrow

[jessfraz]

ill rebase and fix this pr tomorrow

[pierremtb]

@jessfraz Tried to give you a headstart, did git merge main -X theirs and just redo-sim-tests. Feel free to revert if I broke something