Remove duplicate Jenkins push trigger

Jenkinsfile

@@ -79,9 +79,6 @@ pipeline {
     buildDiscarder logRotator(artifactDaysToKeepStr: '', artifactNumToKeepStr: '', daysToKeepStr: '7', numToKeepStr: '10')
     disableConcurrentBuilds()
   }
-  triggers {
-    githubPush()
-  }
   parameters {
     booleanParam(name: 'ANDROID_BUILD', defaultValue: true, description: 'Build an Android version')
     booleanParam(name: 'ANDROID_BUILD_MAESTRO', defaultValue: true, description: 'Build an Android Maestro version')

docs/JENKINS_TRIGGER_VALIDATION.md

@@ -0,0 +1,161 @@
+# Jenkins trigger validation
+
+Date: September 3, 2026
+
+PR: [#6166 — Remove duplicate Jenkins push trigger](https://github.com/EdgeApp/edge-react-gui/pull/6166)
+
+## Result
+
+The proposed change was validated directly in the live
+`EdgeApp/edge-react-gui` multibranch job using the dedicated
+`test-parm` and `test-colby` branches.
+
+| Controlled scenario | Before change | After change |
+|---|---:|---:|
+| Isolated Parm update | Extra Parm build in 4/5 trials | Extra Parm build in 0/5 trials |
+| Colby updated while Parm remained active | Extra Parm build in 10/10 trials | Extra Parm build in 0/10 trials |
+| Colby's own update in the overlap series | Extra Colby build in 5/10 trials | Extra Colby build in 0/10 trials |
+
+Across all 15 post-change trials:
+
+- all 15 Parm updates produced exactly one Parm build;
+- all 10 Colby updates produced exactly one Colby build;
+- all 25 expected builds completed successfully;
+- no extra same-revision build occurred.
+
+The production PR removes only this three-line block:
+
+```groovy
+triggers {
+  githubPush()
+}
+```
+
+GitHub Branch Source remains responsible for normal branch-update delivery.
+
+## Test design
+
+The test used a frozen lightweight Jenkinsfile on the two dedicated branches.
+Parm remained active for a fixed window before checkout. Colby performed a
+short checkout and delay. The harness did not install dependencies, compile the
+application, run application tests, sign, upload, or deploy anything.
+
+The baseline and post-change harness trees differed only by the same
+`githubPush()` block and its adjacent blank line. That is four textual lines
+in the harness, while the production PR itself deletes the three-line block.
+
+Two series were measured before and after the change:
+
+1. **Isolated update:** update Parm once and count all Parm builds for that
+   revision.
+2. **Timed overlap:** update Parm, confirm its expected build is active, then
+   update Colby and count all builds for both controlled revisions.
+
+Every retained overlap was classified after completion using archived Branch
+Source event time, queue-entry time, and build-completion time.
+
+## Primary evidence
+
+### Isolated Parm updates
+
+Before the change, four of five controlled Parm revisions produced two Parm
+builds:
+
+- [#41](https://hq.edge.app/job/edge-react-gui/job/test-parm/41/) and
+  [#42](https://hq.edge.app/job/edge-react-gui/job/test-parm/42/)
+- [#44](https://hq.edge.app/job/edge-react-gui/job/test-parm/44/) and
+  [#45](https://hq.edge.app/job/edge-react-gui/job/test-parm/45/)
+- [#46](https://hq.edge.app/job/edge-react-gui/job/test-parm/46/) and
+  [#47](https://hq.edge.app/job/edge-react-gui/job/test-parm/47/)
+- [#48](https://hq.edge.app/job/edge-react-gui/job/test-parm/48/) and
+  [#49](https://hq.edge.app/job/edge-react-gui/job/test-parm/49/)
+
+The fifth revision built once as
+[#43](https://hq.edge.app/job/edge-react-gui/job/test-parm/43/).
+
+After the change, all five revisions built exactly once as Parm
+[#61](https://hq.edge.app/job/edge-react-gui/job/test-parm/61/) through
+[#65](https://hq.edge.app/job/edge-react-gui/job/test-parm/65/).
+
+### Timed Parm/Colby overlap
+
+All ten retained baseline overlaps produced two Parm builds for the same
+revision:
+
+| Trial | Parm builds | Colby builds |
+|---:|---|---|
+| 1 | [#21](https://hq.edge.app/job/edge-react-gui/job/test-parm/21/), [#22](https://hq.edge.app/job/edge-react-gui/job/test-parm/22/) | [#8](https://hq.edge.app/job/edge-react-gui/job/test-colby/8/), [#9](https://hq.edge.app/job/edge-react-gui/job/test-colby/9/) |
+| 5 | [#29](https://hq.edge.app/job/edge-react-gui/job/test-parm/29/), [#30](https://hq.edge.app/job/edge-react-gui/job/test-parm/30/) | [#16](https://hq.edge.app/job/edge-react-gui/job/test-colby/16/) |
+| 6 | [#31](https://hq.edge.app/job/edge-react-gui/job/test-parm/31/), [#32](https://hq.edge.app/job/edge-react-gui/job/test-parm/32/) | [#17](https://hq.edge.app/job/edge-react-gui/job/test-colby/17/) |
+| 7 | [#33](https://hq.edge.app/job/edge-react-gui/job/test-parm/33/), [#34](https://hq.edge.app/job/edge-react-gui/job/test-parm/34/) | [#18](https://hq.edge.app/job/edge-react-gui/job/test-colby/18/) |
+| 8 | [#35](https://hq.edge.app/job/edge-react-gui/job/test-parm/35/), [#36](https://hq.edge.app/job/edge-react-gui/job/test-parm/36/) | [#19](https://hq.edge.app/job/edge-react-gui/job/test-colby/19/) |
+| 9 | [#37](https://hq.edge.app/job/edge-react-gui/job/test-parm/37/), [#38](https://hq.edge.app/job/edge-react-gui/job/test-parm/38/) | [#20](https://hq.edge.app/job/edge-react-gui/job/test-colby/20/), [#21](https://hq.edge.app/job/edge-react-gui/job/test-colby/21/) |
+| 10 | [#39](https://hq.edge.app/job/edge-react-gui/job/test-parm/39/), [#40](https://hq.edge.app/job/edge-react-gui/job/test-parm/40/) | [#22](https://hq.edge.app/job/edge-react-gui/job/test-colby/22/) |
+| 12 | [#54](https://hq.edge.app/job/edge-react-gui/job/test-parm/54/), [#55](https://hq.edge.app/job/edge-react-gui/job/test-parm/55/) | [#28](https://hq.edge.app/job/edge-react-gui/job/test-colby/28/), [#29](https://hq.edge.app/job/edge-react-gui/job/test-colby/29/) |
+| 13 | [#56](https://hq.edge.app/job/edge-react-gui/job/test-parm/56/), [#57](https://hq.edge.app/job/edge-react-gui/job/test-parm/57/) | [#30](https://hq.edge.app/job/edge-react-gui/job/test-colby/30/), [#31](https://hq.edge.app/job/edge-react-gui/job/test-colby/31/) |
+| 14 | [#58](https://hq.edge.app/job/edge-react-gui/job/test-parm/58/), [#59](https://hq.edge.app/job/edge-react-gui/job/test-parm/59/) | [#32](https://hq.edge.app/job/edge-react-gui/job/test-colby/32/), [#33](https://hq.edge.app/job/edge-react-gui/job/test-colby/33/) |
+
+In the first retained trial, the Colby event arrived while Parm #21 was active.
+Parm #22 entered the queue 7.814 seconds later, before #21 checked out or
+completed. Both Parm builds used revision
+`d11152b006ad85231e5efeaf988f80f02b473d4b`. The expected build recorded the
+Branch Source cause; the extra build recorded only generic push causes.
+
+After the change, all ten overlap trials produced exactly one Parm build
+([#66](https://hq.edge.app/job/edge-react-gui/job/test-parm/66/) through
+[#75](https://hq.edge.app/job/edge-react-gui/job/test-parm/75/)) and one Colby
+build ([#35](https://hq.edge.app/job/edge-react-gui/job/test-colby/35/) through
+[#44](https://hq.edge.app/job/edge-react-gui/job/test-colby/44/)).
+
+In every post-change overlap, the Colby event arrived while Parm remained
+active, with 23.494–27.957 seconds remaining. Each expected build recorded one
+Branch Source event cause and no generic push cause.
+
+## Exclusions and replacements
+
+Baseline overlap attempts 2–4 produced extra Parm builds, but their queue
+entries predated the controlled Colby event. They were reclassified as
+isolated-update observations and excluded from the overlap denominator.
+
+Attempt 11 was excluded for the same reason. Replacement attempts 12–14 met
+the ordering rule, producing the final declared cohort of ten exact overlaps.
+The excluded attempts remain in the
+[complete trial ledger](JENKINS_TRIGGER_VALIDATION_TRIALS.csv).
+
+## Secondary Colby observation
+
+In retained baseline overlap trials 1, 9, 12, 13, and 14, Colby's controlled
+revision also built twice. In every pair, the expected build recorded Branch
+Source and generic causes; the extra build recorded only the generic cause.
+Each extra entered the queue while the expected Colby build remained active.
+
+All ten post-change Colby revisions built exactly once. This is corroborating
+overlap evidence, not an isolated Colby-only test, because Parm was active at
+the same time.
+
+## Consistency checks
+
+An independent read-only review of the retained post-change cohort confirmed:
+
+- all 15 post-change trials were valid;
+- no measured window contained an unrelated branch event;
+- archived event records were continuous with no rollover;
+- child configurations stayed in the expected state throughout each phase;
+- the Parm polling-log timestamp did not change in any post-change trial;
+- Git revisions agreed across commit, event, update, and Jenkins build records;
+- all expected post-change builds succeeded.
+
+The verified manifest covers 627 evidence files (14 MB), including trial
+records, Jenkins build metadata, logs, configuration snapshots, and event
+snapshots. The archive and its manifest are retained separately rather than
+committed to the application repository. The manifest's SHA-256 is
+`0dfb815226d3f076aa7b67064ddd247a7c9ba15b1b4cc4fc68c251ad9e0cd7ba`.
+
+## Scope
+
+These results directly cover the isolated-update and timed-overlap scheduling
+paths exercised by this validation. They do not claim to eliminate every
+possible source of repeated Jenkins requests.
+
+The validation-branch state after measurement does not affect the PR branch or
+its merge.