[Affected versions]: Fx50 [Affected platforms]: Windows 10 x64 Mac OS X 10.11 Ubuntu 14.04 LTS [Steps to reproduce]: 1. Launch Firefox. 2. Go to "about:config" and make sure the pref: "toolkit.telemetry.eneblet" is set on "true". 3. Go to " about:preferences#search" and set keywords for the aliased search engines( i.e. "y" for yahoo, "w" for wikipedia). 4. Go to :about:telemetry" and check the "Raw JSON" checkbox. 5. Open a new tab and serch a string with an aliased search engine using the keyword from step 3. 6. Refresh "about:telemetry" and search for "search_counts"(using Ctrl/Cmd + F will simplify the process). [Expected result]: The SEARCH_COUNTS creates buckets for every search engine used (i.e "yahoo.urlbar" for yahoo). [Actual result]: The aliased search engines searches are not accounted. [Regression Range]: Last bad: 20161013030204 First good: 20161014060324 Pushlog: https://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=0c899672fff6ae00f5b3affbec48ee4daac35fa1&tochange=f03e2740d604d339ed553dad62a3fc54c317f8fa [Additional notes]: a. This is reproducible only on Fx50 this has been fixed on nightly and aurora by bug 1306639. b. The bucket for the default search engine is created and accounted. c. Searches via the Search Bar are accounted.
it may end up being a wontfix considered the code freeze, I'll see how hard it would be to fix it, and eventually check for a workaround for our shield study. That said, we don't have aliases by default and it's a not so much used feature in the wild, so it should not skew our data in a significant way.
Javaun agrees that we shouldn't be terribly concerned about the skew in numbers we will get from the SHIELD study. Feel free to wontfix if it is anything more than a trivial workaround.
I'm fixing this add-on side, it's easy enough.