In about:preferences#privacy, - Check "Allow Nightly to send technical and interaction data to Mozilla" (datareporting.healthreport.uploadEnabled=true) - Uncheck "Allow Nightly to install and run studies" (app.shield.optoutstudies.enabled=false) Then go to about:studies. A message says > Nightly may install and run studies from time to time I need to disable "Allow Nightly to send technical and interaction data to Mozilla" in order to get "No new studies will run." This is very confusing. If app.shield.optoutstudies.enabled=false alone doesn't prevent studies, then "Allow Nightly to install and run studies" in about:preferences should be changed to explain whatever the pref really does. Otherwise about:studies should also check app.shield.optoutstudies.enabled instead of just datareporting.healthreport.uploadEnabled. This started in https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=8fb0418948834f071dd99f6b8256b746b4a8adce&tochange=94f5c92fc711a1e1e76ea54d52482cbbd2db0401

Thanks for the bug report. I was able to duplicate what you're seeing as well. I'm skeptical that the bug started in the changeset you linked, since that commit tends to break history views. I suspect it was pre-existing when we moved all the code around. How did you narrow it down to that changeset? I'm fairly sure, but haven't verified yet, that we are honoring the preferences in the actual behavior of installing add-ons, and the bug is only related to about:studies. I will also double check about that though.

(In reply to Michael Cooper [:mythmon] from comment #1) > I'm skeptical that the bug started in the changeset you linked, since that > commit tends to break history views. I suspect it was pre-existing when we > moved all the code around. How did you narrow it down to that changeset? I used mozregression-gui. Before that changeset, the message box was empty. After the changeset, I started getting "Nightly may install and run studies from time to time" despite app.shield.optoutstudies.enabled=false

[Tracking Requested - why for this release]: we don't want this to get bigger than it already is; should be resolved ASAP.

I did more research here, and the problem happens happens specifically when no studies have ever run on the profile. When this is the case, the message chosen doesn't vary based on if studies are enabled or not. To be clear, the pref is respected: when the preference is changed studies are not installed and existing studies are ended. The bug here is the wrong message being shown only.

Pushed by mcooper@mozilla.com: https://hg.mozilla.org/integration/autoland/rev/f0ed99b29aaf Reflect study opt-out in about:studies when Normandy is generally enabled. r=Gijs

Backed out 2 changesets (bug 1501877, bug 1498940) for failing browser_about_studies.js push that caused the backout: https://treeherder.mozilla.org/#/jobs?repo=autoland&resultStatus=testfailed%2Cbusted%2Cexception%2Cretry%2Cusercancel%2Crunnable&revision=f0ed99b29aafb5fd7f613a000634b53927af4259&selectedJob=207819106 failures: https://treeherder.mozilla.org/#/jobs?repo=autoland&resultStatus=testfailed%2Cbusted%2Cexception%2Cretry%2Cusercancel%2Crunnable&revision=f0ed99b29aafb5fd7f613a000634b53927af4259&selectedJob=207819106&searchStr=linux%2Cx64%2Cdebug%2Cmochitests%2Cwith%2Ce10s%2Ctest-linux64%2Fdebug-mochitest-browser-chrome-e10s-4%2Cm-e10s%28bc4%29 backout: https://hg.mozilla.org/integration/autoland/rev/d08c28ed42f1bef0b77d38064759b797f377a998

Pushed by mcooper@mozilla.com: https://hg.mozilla.org/integration/autoland/rev/9de1bee6adc2 Reflect study opt-out in about:studies when Normandy is generally enabled. r=Gijs

Please nominate this for Beta approval when you get a chance.

Comment on attachment 9019889 [details] Bug 1498940 - Reflect study opt-out in about:studies when Normandy is generally enabled. r=Gijs [Beta/Release Uplift Approval Request] Feature/Bug causing the regression: None User impact if declined: When a user has disabled studies but not Normandy, about:studies incorrectly shows that studies may run. Is this code covered by automated tests?: Yes Has the fix been verified in Nightly?: No Needs manual test from QE?: No If yes, steps to reproduce: List of other uplifts needed: None Risk to taking this patch: Low Why is the change risky/not risky? (and alternatives if risky): Only a change to the display logic of about:studies, not any of the actual logic of running studies or Normandy String changes made/needed:

This has conflicts with Beta - do we need bug 1501877 also?

Oops. Yes, we need bug 1501877. We'll also need bug 1502169 to avoid some intermittent test failures. The three patches were developed together and split to separate concerns in review.

Comment on attachment 9019889 [details] Bug 1498940 - Reflect study opt-out in about:studies when Normandy is generally enabled. r=Gijs [Triage Comment] Fixes a confusing UI issue, approved for 64.0b6.

Reproduced the issue on Fx65 29-10-2018 / verified as fixed on Nightly 31-10-2018 on Windows 10/Windows 8.1. Given the nature of the fix, I don't think a cross platform verification is required. With the fix, when "Allow Nightly to install and run studies" is checked off, the about:studies lists: "This is a list of studies that you have participated in. No new studies will run."

Verified as fixed on 64.0b6 2018-11-01 on OSX 10.13/ Windows 10 / Ubuntu 16.04.