You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Recommendation: no regression penalty. Every linked deploy blocker was demoted (none actually blocked a deploy) because the Copy Settings feature is gated behind the BULK_EDIT_WORKSPACES beta and never reached production users — which is exactly what the beta is for.
@MelvinBot the above is partially true. We don't want any regressions because, even if they're on staging or in beta, the still require time and resources. I think a better litmus test to decide if the regression should lower the price of a job is "should the contributor and/or C+ have caught the regressions" With features being new it's more likely that, because it's new, it would be more difficult for teh C+ to have caught the regression (if at all possible).
🤖 Fair — I'll drop "behind a beta" as the deciding factor. Staging/beta regressions still burn CME, Applause, and C+ time, so the right question is your litmus test: should the contributor and/or C+ have reasonably caught each one? Re-scoring the four on that basis:
Backend root cause. The server reconstructs the rates; the App PR only sends parts and never touches rate data (CopyPolicySettings.ts:317-323), so neither code review nor a normal App-side test would surface it. Durable fix is backend.
The one genuine App-side gap. The copy flow copied approval chains but not the default approver, producing a spurious workflow entry — that logic lives in reviewable App code (CopyPolicySettings.ts). Arguably catchable, but a subtle omission on a brand-new feature.
Judgment call
Bottom line under your test: at most one of the four (#92332) is a "should-have-caught" App regression. The other three are not a bug, pre-existing, or backend root-cause — and as you note, catchability is lower on a brand-new feature.
Payment options for C+ shubham1206agra (base $250; fedirjh is a contractor, no payment):
My recommendation: penalize at most #92332 (→ $125); lean to no penalty if you weight the new-feature difficulty heavily. #92329 / #92385 / #92397 shouldn't reduce payment.
Activity
mitarachim commentedon Jun 2, 2026
Deploy Blocker #92329 was identified to be related to the PR #91241 that attempted to fix this issue.
mitarachim commentedon Jun 2, 2026
Deploy Blocker #92332 was identified to be related to the PR #91241 that attempted to fix this issue.
melvin-bot commentedon Jun 2, 2026
If you are the assigned CME please investigate whether the linked PR caused a regression and leave a comment with the results.
If a regression has occurred and you are the assigned CM follow the instructions here.
If this regression could have been avoided please consider also proposing a recommendation to the PR checklist so that we can avoid it in the future.
melvin-bot commentedon Jun 2, 2026
If you are the assigned CME please investigate whether the linked PR caused a regression and leave a comment with the results.
If a regression has occurred and you are the assigned CM follow the instructions here.
If this regression could have been avoided please consider also proposing a recommendation to the PR checklist so that we can avoid it in the future.
25 remaining items
shubham1206agra commentedon Jun 29, 2026
@mallenexpensify Bump here
mallenexpensify commentedon Jul 1, 2026
@MelvinBot the above is partially true. We don't want any regressions because, even if they're on staging or in beta, the still require time and resources. I think a better litmus test to decide if the regression should lower the price of a job is "should the contributor and/or C+ have caught the regressions" With features being new it's more likely that, because it's new, it would be more difficult for teh C+ to have caught the regression (if at all possible).
mallenexpensify commentedon Jul 1, 2026
@shubham1206agra please address the deploy blockers and fill out the BZ checklist once Melv has completed their review
MelvinBot commentedon Jul 1, 2026
🤖 Fair — I'll drop "behind a beta" as the deciding factor. Staging/beta regressions still burn CME, Applause, and C+ time, so the right question is your litmus test: should the contributor and/or C+ have reasonably caught each one? Re-scoring the four on that basis:
fedirjh/ishpaul777) agreed it wasn't a bug; re-scoped as enhancement (#91792, its own PR + payment).partsand never touches rate data (CopyPolicySettings.ts:317-323), so neither code review nor a normal App-side test would surface it. Durable fix is backend.CopyPolicySettings.ts). Arguably catchable, but a subtle omission on a brand-new feature.Bottom line under your test: at most one of the four (#92332) is a "should-have-caught" App regression. The other three are not a bug, pre-existing, or backend root-cause — and as you note, catchability is lower on a brand-new feature.
Payment options for C+
shubham1206agra(base $250;fedirjhis a contractor, no payment):My recommendation: penalize at most #92332 (→ $125); lean to no penalty if you weight the new-feature difficulty heavily. #92329 / #92385 / #92397 shouldn't reduce payment.
Final call is yours on the checklist.
shubham1206agra commentedon Jul 2, 2026
@mallenexpensify #92332 is also BE issue. So, there should be no penalty here
shubham1206agra commentedon Jul 2, 2026
Also, checklist is not required here as it should be covered under docs
mallenexpensify commentedon Jul 2, 2026
Payment Summary
Contributor+: @shubham1206agra due $250 via NewDot
JmillsExpensify commentedon Jul 12, 2026
$250 approved for @shubham1206agra