Page MenuHomePhabricator

Remove (redundant) 'mark as read' function from notifications dotdotdot menu
Closed, ResolvedPublic

Description

Now that the blue dot for mark as read/unread is in the corner of every notification, it's redundant in the dotdotdot menu.

In the vast majority of notifications, mark as unread is currently the only function in the dotdotdot menu. So removing it enables us to remove a lot of clutter. Just to spell out the details with regard to that:

  • In the very few cases where there are items in the dotdotdot menu aside from mark as unread/read, the menu should still display.
  • In all other cases (where mark as read/unread is the only option), the menu should no longer display.
  • Ideally, we'll construct this menu so that if we add new options to a given notification type -- I'm thinking of a secondary link beyond the two currently allowed in the main message display -- those options will automatically cause the menu to display. (Or maybe that already works -- see T125949)

Event Timeline

Restricted Application added subscribers: Zppix, Aklapper. · View Herald TranscriptJul 15 2016, 9:42 PM
jmatazzoni updated the task description. (Show Details)

Have you done any research on how many people actually know that the blue dot is clickable and means "mark as read"? I've seen many people using either the dotdotdot menu or the button "mark group as read"... Especially newbies tend to go to places, where is verbal command available.

Have you done any research on how many people actually know that the blue dot is clickable and means "mark as read"?

Hi @Danny_B, yes, we did usability research on this very point, and users either knew what the blue dot was allready (very likely from programs like Mac Mail or Facebook, which offer identical or similar tools) or figured it out easily. Discovering what the dot does/means is pretty effortless because it's so visible and changes so obviously and is correlated so consistently with other signals of read/unread-ness (e.g., background color). Also the dots display tooltips on hover.

So overall we're pretty confident this is a very positive change. Thanks for asking!

Can I see the data, please? I mean doing research between (presumably American) people who use Macs and/or Facebook doesn't reflect experiences of people who don't. Again, I've seen many people to simply ignore the blue dot (considered it just as a status indicator or some even as plain design thingy).

The research on the Notification page was one of the studies where we presented alternatives to the "X" (we used three versions of the same prototype: original, blue dot and checkmark) and all worked well.

In this context getting new notifications and comparing how those are represented in contrast to the old ones (or the current one after you clicked on it to read it) helps to reinforce the understanding on the visual cues.

The use of a bright dot to signal new activity is not uncommon. It is also present in Thunderbird and other messaging-related software.

Seriously are 3-5 editors sample of the entire population? What is the further breakdown of them? That's the data I was actually asking for. Age, newbie/poweruser, experiences with other Mac, Facebook... I can immediatelly provide a counter-research with the same number of people giving the totally opposite result.

For the record: I personally have no issue about the blue dot, I just care about others for whome the dot is not straightforward.

Seriously are 3-5 editors sample of the entire population?

This is qualitative research, and this is a pretty typical size for this kind of thing.

It's not a survey with sample sizes, etc. If we were trying a radically new interface, we might need an A/B test. But we're basically adopting a standard UI metaphor.

Of course, we will keep listening about this (and the product in general)..

I can immediatelly provide a counter-research with the same number of people giving the totally opposite result.

Cherry-picking people based on a pre-determined outcome is not how qualitative research (or any research) works.

I can immediatelly provide a counter-research with the same number of people giving the totally opposite result.

Cherry-picking people based on a pre-determined outcome is not how qualitative research (or any research) works.

AGF. Besides it was rather rhetorical. But if I was doing the same research in paralell to yours, I would get opposite results. No need for cherrypicking. It's just the area where you ask. I am still missing the data actually. It's easy to get that result, if 5 out of 5 tested people are Mac users and Facebookers... If you will ask people in Versace shop, how are they satisfied with their level of living, you will most probably get 100% of positive answers. And vice versa, if you asked in some homeless shelter...

Change 301718 had a related patch set uploaded (by Mooeypoo):
Remove 'mark as unread' from dotdotdot menu in items

https://gerrit.wikimedia.org/r/301718

Change 301718 merged by jenkins-bot:
Remove 'mark as unread' from dotdotdot menu in items

https://gerrit.wikimedia.org/r/301718

Checked the fix in 1.28.0-alpha (86ffeb1) - redundant 'Mark as read/unread' is removed from the dotdotdot menu.

jmatazzoni closed this task as Resolved.Jul 29 2016, 11:58 PM