CLA Signature Action: All authors have signed the CLA. You may need to manually re-run the blocking PR check if it doesn't pass in a few minutes.

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (3844eac) 68.47% compared to head (babc8df) 68.47%.
Report is 3 commits behind head on develop.

@@           Coverage Diff            @@
##           develop   #22762   +/-   ##
========================================
  Coverage    68.47%   68.47%           
========================================
  Files         1088     1088           
  Lines        42956    42956           
  Branches     11436    11436           
========================================
  Hits         29410    29410           
  Misses       13546    13546           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

1. Let's centralize those selectors, you're repeating some of the same ones in 3 places already. Create /test/e2e/helpers/selectors.ts and import from there.
2. This is an optional challenge, but it would be nice to write new tests in TypeScript. We're supposed to be writing all new things in TypeScript, but few people are following it. I wrote some samples of TS E2E in the /test/e2e/accounts folder.

@HowardBraham ,Thank you for your keen observation regarding centralizing the object selectors. I absolutely agree that there is duplication in that regard. Mariona is currently working on implementing the Page Object Model, which will be a significant framework change, and we will address this issue as part of that process.

I did explore TypeScript briefly, using your accounts as a reference, but I encountered some challenges. I plan to allocate some time today to give it another try. I agree with your recommendation that we should transition to writing our code in TypeScript.

I was thinking about what Mark Stacey said today with our very large CircleCI usage. In that context, it makes sense to combine and form larger tests. I tried, and this could be condensed into a single test that takes 55 seconds to run. This is true because nothing has to be "reset" in between tests.

How do other people feel about this?

I was thinking about what Mark Stacey said today with our very large CircleCI usage. In that context, it makes sense to combine and form larger tests. I tried, and this could be condensed into a single test that takes 55 seconds to run. This is true because nothing has to be "reset" in between tests.

How do other people feel about this?

Hi @HowardBraham , Yes, I always agree with the idea of smart coupling for e2e test, It reduces execution time, it's more "end-to-end" to test a user flow, it's closer to real user's behaviour, as real user don't have a clean setup each time. Additionally, as you mentioned, it saves money on CircleCI. We do acknowledge the downside that if one step fails, the whole test will fail, making it slightly more challenging to debug. However, overall, the advantages outweigh the disadvantages, making it reasonable to combine certain our e2e tests.