Bugzilla

Cool thanks for the update, I'll work on the revision restricting the feature to nightly builds and adding telemetry. I'm unsure about adding pings in both places, because I think it means there would be twice as much data from users who have auto update enabled as from users who have it disabled? And this feature is likely more desirable for users who update manually but just want badges/banners rather than doorhangers. I think the main workaround (aside from adding telemetry to AppUpdater.jsm) is if I were to add the ping at the very top of _checkForBackgroundUpdates, `if (this.manualUpdateOnly) ...` before it returns, but I'm not sure if that would run afoul of bug 1599590

Cool thanks for the update, I'll work on the revision restricting the feature to nightly builds and adding telemetry. I'm unsure about adding pings in both places, because I think it means there would be twice as much data from users who have auto update enabled as from users who have it disabled? And this feature is likely more desirable for users who update manually but just want badges/banners rather than doorhangers. I think the main workaround (aside from adding telemetry to AppUpdater.jsm) is if I were to add the ping at the very top of _checkForBackgroundUpdates, `if (this.manualUpdateOnly) ...` before it returns, but I'm not sure if that would run afoul of bug 1599590

I could just make a scalar key here, it seems appropriate and would also avoid that problem, but almost everything else here is in histograms. If there's a particular reason for that I don't wanna screw anything up

Cool thanks for the update, I'll work on the revision restricting the feature to nightly builds and adding telemetry. I'm unsure about adding pings in both places, because I think it means there would be twice as much data from users who have auto update enabled as from users who have it disabled? And this feature is likely more desirable for users who update manually but just want badges/banners rather than doorhangers. I think the main workaround (aside from adding telemetry to AppUpdater.jsm) is if I were to add the ping at the very top of _checkForBackgroundUpdates, `if (this.manualUpdateOnly) ...` before it returns, but I'm not sure if that would run afoul of bug 1599590

I could just make a scalar here, it seems appropriate and would also avoid that problem, but almost everything else here is in histograms. If there's a particular reason for that I don't wanna screw anything up.

Edit: For now I'm just going with a scalar that probes the inverse of the pref value to get a figure of users changing the pref. But if histogram is important lmk and I'll change it.

Cool thanks for the update, I'll work on the revision restricting the feature to nightly builds and adding telemetry. I'm unsure about adding pings in both places, because I think it means there would be twice as much data from users who have auto update enabled as from users who have it disabled? And this feature is likely more desirable for users who update manually but just want badges/banners rather than doorhangers. I think the main workaround (aside from adding telemetry to AppUpdater.jsm) is if I were to add the ping at the very top of _checkForBackgroundUpdates, `if (this.manualUpdateOnly) ...` before it returns, but I'm not sure if that would run afoul of bug 1599590

I could just make a scalar here, it seems appropriate and would also avoid that problem, but almost everything else here is in histograms. If there's a particular reason for that I don't wanna screw anything up.

Edit: For now I'm just going with a boolean scalar that probes the inverse of the pref value to get a figure of users changing the pref. But if histogram is important lmk and I'll change it.

Cool thanks for the update, I'll work on the revision restricting the feature to nightly builds and adding telemetry. I'm unsure about adding pings in both places, because I think it means there would be twice as much data from users who have auto update enabled as from users who have it disabled? And this feature is likely more desirable for users who update manually but just want badges/banners rather than doorhangers. I think the main workaround (aside from adding telemetry to AppUpdater.jsm) is if I were to add the ping at the very top of _checkForBackgroundUpdates, `if (this.manualUpdateOnly) ...` before it returns, but I'm not sure if that would run afoul of bug 1599590

I could just make a scalar here, it seems appropriate and would also avoid that problem, but almost everything else here is in histograms. If there's a particular reason for that I don't wanna screw anything up.

Edit: For now I'm just going with a boolean scalar that probes the inverse of the pref value to get a figure of users changing the pref. But if histogram is important lmk and I'll change it.

Is our goal to record and compare how often a user downloads an update with the pref enabled, how often they install an update with the pref enabled, and how often they do so with the pref disabled, to see whether the pref has a negative impact on users' update habits? I suppose it could also be set up to record the current firefox version to determine whether users of the pref are systematically falling behind on updates. I'm not sure I'm the best person to set that up though, I don't know all the ins and outs of telemetry.

Cool thanks for the update, I'll work on the revision restricting the feature to nightly builds and adding telemetry. I'm unsure about adding pings in both places, because I think it means there would be twice as much data from users who have auto update enabled as from users who have it disabled? And this feature is likely more desirable for users who update manually but just want badges/banners rather than doorhangers. I think the main workaround (aside from adding telemetry to AppUpdater.jsm) is if I were to add the ping at the very top of _checkForBackgroundUpdates, `if (this.manualUpdateOnly) ...` before it returns, but I'm not sure if that would run afoul of bug 1599590

Edit: Is our goal to record and compare how often a user downloads an update with the pref enabled, how often they install an update with the pref enabled, and how often they do so with the pref disabled, to see whether the pref has a negative impact on users' update habits? I suppose it could also be set up to record the current firefox version to determine whether users of the pref are systematically falling behind on updates. I'm not sure I'm the best person to set that up though, I don't know all the ins and outs of telemetry.

I ended up following the histogram format used elsewhere in UpdateService.jsm since I don't fully understand the external/notify/subsequent schema. So I believe it's submitting on every update flow when the user has the pref set to false on a compatible platform & build.