Clicking on the hovercards enabled link element (not the hovercard itself) is sending null as the total interaction time. This affects the test group (Hovercards ON), and not the control group (Hovercards OFF).
Description
Description
Details
Details
| Subject | Repo | Branch | Lines +/- | |
|---|---|---|---|---|
| Regression: correctly log the click action total interaction time | mediawiki/extensions/Popups | master | +4 -7 |
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Resolved | None | T145665 Verify interaction times are logged correctly with hovercards off and on (with the same actions) | |||
| Resolved | • Jhernandez | T146934 Click event is not correctly logging the total interaction time |
Event Timeline
Comment Actions
Change 313258 had a related patch set uploaded (by Bmansurov):
Regression: correctly log the click action total interaction time
Comment Actions
So to clarify, does this issue only affect the test group (Hovercards ON) or also the control group of the Popups schema?
Comment Actions
@Jdlrobson can you review https://gerrit.wikimedia.org/r/#/c/313258/ so that it rides the train tomorrow?
Comment Actions
I only just got a chance to look but I can't verify the fix. I'm finishing up for the day now. @phuedx can you take a look?
Comment Actions
Change 313258 merged by jenkins-bot:
Regression: correctly log the click action total interaction time
Comment Actions
From my review of rEPOP3c0ea0f9cda9: Regression: correctly log the click action total interaction time:
I was able to reproduce the issue with the following steps:
- hover over a link and wait for a hovercard to appear
- open the link in a new tab
- observe the totalInteractionTime property of the opened in a new tab event is null
Comment Actions
Verified in reading-web-staging.wmflabs.org with @phuedx instructions
- hover over a link and wait for a hovercard to appear
- open the link in a new tab
- observe the totalInteractionTime property of the opened in a new tab event is
null
totalInteractionTime is correctly a number, in my case around 3000 for the 3s of interaction time on my tests.