Skip to content

docs(kaneai): correct what several configurations do to Adaptive Heal and Dynamic Test - #3516

Open
anubhavsinghmaar wants to merge 1 commit into
LambdaTest:stagefrom
anubhavsinghmaar:docs/kaneai-healing-one-draft
Open

anubhavsinghmaar wants to merge 1 commit into
LambdaTest:stagefrom
anubhavsinghmaar:docs/kaneai-healing-one-draft

Conversation

@anubhavsinghmaar

Copy link
Copy Markdown
Contributor

Corrects a factual error introduced in #3492 and carried into #3515, both now merged.

What the page says today

Only the first test instance picked up for execution is eligible. The remaining instances in that run replay their recorded steps and are not repaired or re-authored. Running a single test case across several Chrome versions or several operating system versions is not supported for these strategies.

If you need every combination covered, run each version as its own test run.

Why that is wrong

It describes eligibility, and the real behaviour concerns the draft. As written it leads a reader to two wrong conclusions: that most of their test instances will be silently skipped, and that they should split one run into several to get coverage.

What it says now

The claim is removed from When these strategies do not apply, where it did not belong, because this is not a case of the strategies failing to apply.

The correct behaviour is stated as a note in Reviewing a change, beside the version it is actually about:

One test case, one draft
A test run can include the same test case under several configurations, such as more than one Chrome version or more than one operating system. It still produces a single draft, because it is one test case.

Verification

🤖 Generated with Claude Code

… and Dynamic Test

The page said only the first test instance picked up for execution is eligible, that the rest
replay their recorded steps without being repaired, and that a test case across several Chrome
or operating system versions is not supported. That is not the behaviour, and it sent a reader
to the wrong conclusion twice: that most of their instances would be skipped, and that they
should split a run to get coverage.

What actually happens concerns the draft, not eligibility. A test run can include the same test
case under several configurations, and it produces a single draft, because it is one test case.

The claim is removed from When these strategies do not apply, where it did not belong, since
this is not a case of the strategies failing to apply. The correct behaviour is stated as a note
in Reviewing a change, beside the version it is about.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant