Page MenuHomePhabricator

CU 2.0: Filter out exact UAs in the Compare tab
Open, MediumPublic

Description

Split from T240530: CU 2.0: Filtering from the data table in Compare tab. See T240530#6011630 for more context.

Goal

In the table or lists of results on the CU special page, the user should be able to filter out data using the delete icon that shows up when they hover over a cell.

Mock

https://prtksxna.github.io/wmf-cu-prototype/compare.html (hover over Chrome 65 in the first row)

Unhighlighted row
Screenshot 2019-12-12 at 12.59.33 PM.png (218×1 px, 50 KB)
Hovering over row
Screenshot 2019-12-12 at 1.01.22 PM.png (222×1 px, 55 KB)
Hovering over remove icon (highlights the rows that'll get removed in red)
Screenshot 2019-12-12 at 12.59.44 PM.png (730×1 px, 180 KB)

Acceptance criteria:

  • Requisite filters:
    • Filter out user-agent
  • After the user clicks the cross button, the exact user-agent is added to the filters widget

image.png (227×775 px, 15 KB)

  • Filters persist across pagination
  • When a new filter is added by clicking the icon, pagination returns to the first page of results.
  • Clicking is a submission of the filter and will refresh the page.
Note

Event Timeline

Niharika triaged this task as Medium priority.Apr 3 2020, 9:16 PM
Niharika created this task.
Niharika renamed this task from CU 2.0: Filter out UAs in the Compare tab to CU 2.0: Filter out exact UAs in the Compare tab.Apr 8 2020, 4:43 PM
Niharika updated the task description. (Show Details)

In {T248588}, we found that the Compare query is already pushing the acceptable time limit, and that filters increase the time. Adding user agent filters will increase that problem further.

If we decide this is a really important feature, we might need to rethink how we get the data. E.g. one idea that has been floated before is adding another table so that we don't have to aggregate results as the query is run. Solutions like this would take some work but could improve things in the long-run.

E.g. one idea that has been floated before is adding another table so that we don't have to aggregate results as the query is run. Solutions like this would take some work but could improve things in the long-run.

I'm supposed to make a task where we can collect our thoughts on this topic. I'll do that today and link it here.