NEC-111: e2e flake: collaborator pointer test sees a pointer after a touch move #173

Merged

nectenda-agent merged 3 commits from worktree-nec-111-e2e-flake-collaborator-pointer-test-sees into main 2026-09-25 17:27:00 +01:00

Conversation 0 Commits 3 Files changed 3 +377 -2

nectenda-agent commented 2026-09-25 16:53:06 +01:00

Collaborator

Copy link

Task: NEC-111 — https://projectron.nerchure.com/tasks/111

Fixes the intermittent 'a touch drew a pointer' failure in the collaborator-pointer e2e test (CI tasks 1320, 1344). Both failures drew the same fresh pointer from B at the same pixel: something pointing from a fixed place, not the touch. One trusted OS mouse move into B during the check reproduces the failure exactly (collab-b, edge null, opacity 0.72, animationDelay 2.9998s against CI's 2.9999s). The pointer tests now stop trusted pointer events at the window, clear B's pointer before the touch, and label each test step. Every run prints the trusted events it saw, with screen position, focus, whether each was blocked, and the step, so CI can confirm or rule out the Xvfb stationary-cursor explanation. A new test proves both halves: an unshielded OS move is drawn in A, a shielded one is recorded and not drawn. Mutation-checked: a touch-sending plugin fails the touch check, and a shield that doesn't block fails the new test. Test-only change. Filed along the way: NEC-112 (pointer returning to the same anchor drawn already faded) and NEC-113 (directory-manifest ETag hashes volatile fields, flaky 304 test).

Spec: docs/changes/NEC-111-pointer-touch-flake/spec.md

Changelog

NONE

Task: NEC-111 — https://projectron.nerchure.com/tasks/111 Fixes the intermittent 'a touch drew a pointer' failure in the collaborator-pointer e2e test (CI tasks 1320, 1344). Both failures drew the same fresh pointer from B at the same pixel: something pointing from a fixed place, not the touch. One trusted OS mouse move into B during the check reproduces the failure exactly (collab-b, edge null, opacity 0.72, animationDelay 2.9998s against CI's 2.9999s). The pointer tests now stop trusted pointer events at the window, clear B's pointer before the touch, and label each test step. Every run prints the trusted events it saw, with screen position, focus, whether each was blocked, and the step, so CI can confirm or rule out the Xvfb stationary-cursor explanation. A new test proves both halves: an unshielded OS move is drawn in A, a shielded one is recorded and not drawn. Mutation-checked: a touch-sending plugin fails the touch check, and a shield that doesn't block fails the new test. Test-only change. Filed along the way: NEC-112 (pointer returning to the same anchor drawn already faded) and NEC-113 (directory-manifest ETag hashes volatile fields, flaky 304 test). Spec: `docs/changes/NEC-111-pointer-touch-flake/spec.md` ## Changelog NONE

nectenda-agent added the

area:plugin

label 2026-09-25 16:53:06 +01:00

nectenda-agent added 3 commits 2026-09-25 16:53:06 +01:00

Keep the OS cursor out of the pointer tests, and record what it sends ... 9b0cf23442

The touch check in the collaborator-pointer tests failed on CI twice with
a fresh pointer from B at the same pixel both times. A trusted mouse move
into B reproduces that failure exactly. So the pointer tests now stop
trusted pointer events at the window. They record each one with where it
landed, whether it was blocked, and the step running, and print the log on
every run. They also clear B's pointer before the touch. A new test shows
both halves: an unshielded OS move is drawn in A, a shielded one is not.

Task: NEC-111

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FGckKvM58WJBmZm5GHAs6h

Mark the pointer-event log as a deliberate console line, as the probes do ... e319dcbbd9

Task: NEC-111

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FGckKvM58WJBmZm5GHAs6h

Label every pointer test's step, and correct two claims the review caught ... 40b63fcad5

The tests after the OS cursor proof inherited its step label, so events
they cause would have read as the proof test's. The NEC-112 comment named
the wrong half, and the spec described a CDP fallback that was not built.

Task: NEC-111

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FGckKvM58WJBmZm5GHAs6h

cruelacid force-pushed worktree-nec-111-e2e-flake-collaborator-pointer-test-sees from 40b63fcad5 to b368b34172 2026-09-25 17:20:59 +01:00 Compare

nectenda-agent merged commit b368b34172 into main 2026-09-25 17:27:00 +01:00

Sign in to join this conversation.

Reviewers

No reviewers

Labels

Clear labels   

area:docs

docs/

  

area:identity

packages/identity

  

area:ops

deploy/, CI, hosting

  

area:plugin

packages/plugin

  

area:server

packages/server

  

channel:community

Obsidian forum, Obsidian Discord, Reddit

  

channel:direct

Cold outreach to named firms

  

channel:owned

nectenda.com pages, the feedback board

  

channel:press

Hacker News, newsletters, outlets, YouTube creators

  

channel:social

Bluesky, Mastodon, X, YouTube accounts

  

e2ee-constrained

End-to-end encryption changes the shape of the answer

  

gate:at-ga

Held until Gates A-C close and terms/privacy leave DRAFT

  

gate:pre-ga

Safe to do before the launch gates close

  

marketing

Reach and positioning; work published off nectenda.com

  

parity

Relay feature-parity programme

  

relay:absent

4 - Not supported, not requested, not planned at Relay

  

relay:planned

2 - On Relay's roadmap (Active/Planned) or has an open PR

  

relay:requested

3 - Asked for by Relay users or under 'Considering'; not committed

  

relay:supported

1 - Relay ships this today

  

risk:additive

New message type, table or column; backward compatible

  

risk:contract

Changes something existing accounts or stored data depend on

  

risk:none

Client-only; the server is untouched

  

usability

How the product is understood and learned, not what it does

No labels

area:docs

area:identity

area:ops

area:plugin

area:server

channel:community

channel:direct

channel:owned

channel:press

channel:social

e2ee-constrained

gate:at-ga

gate:pre-ga

marketing

parity

relay:absent

relay:planned

relay:requested

relay:supported

risk:additive

risk:contract

risk:none

usability

Milestone

Clear milestone

No items

No milestone

Projects

Clear projects

No items

No project

Assignees

Clear assignees

No assignees

2 participants

Notifications

Subscribe

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set

Reference

Nectenda/nectenda!173

No description provided.