intrigeri:
> (I don't mean to restart this conversation but instead to get some
> kind of closure; I'm happy to come back to it later.)
I meant to provide a closure already in my previous email by creating
#17077 and writing:
-----8<-----
I really don't want to do the change before using it in usability tests.
I could reason about its usability in the abstract but, since it's
already a working piece of code, it might be more efficient to see how
it perform with users.
I will agree if we can confirm that the UX of the Activities overview is
actually better that this brute-force (there's a better UX term this
behavior pattern but I can't remember it...).
A possible next step would be to have a branch that replaces the
Applications menu with the Activities overview so I can test it at some
point while I'm testing other things (I don't think that it deserves a
run of tests on its own).
I created #17077 to track this. A possible next step could be to write a
branch that does the replacement so I could test it at some point.
----->8-----
I don't want to be doing guesswork usability when touching such a core
feature of our OS. Watching GUADEC videos from this year, it's pretty
clear to me that GNOME has been doing guesswork usability themselves
here so, unfortunately, we cannot rely on them too much (and I hope that
we'll be able to help them instead).
For example, it might be something easy to test if we get Sponsor 3 and
this would happen before Debian Bullseye.
But it seems important for you to continue discussing this so I'll
answer you once more.
> sajolida:
>> intrigeri:
>> Now, we have no idea about the usability of the activities overview vs
>> the Applications menu one you're in it. We know that the Applications
>> menu has slightly confusing categories but that people make it quite
>> all-right (I've seen it many many times in usability tests). But we
>> can't say the same of the activities overview as we've never tested it.
>
> While I understand it's not the same as proper usability testing
> focused on exactly this topic, here are my 2cts:
>
> - In another life of mine, several times a year I see 3-9 people
> interact with GNOME for the first time. Almost all of them discover
> the Activities overveiw by themselves. All the apps they need in
> this context are in the dock (favorites) so I don't know how they
> would find other apps.
>
> - According to my notes, starting apps was not a problem during the
> usability testing of Debian Stretch that some of us have run 2.5
> years ago.
The last person I interviewed was using Tails every week for his job
(I haven't asked since when). He complained about a quite typical
Steering Law problem [1] when moving through the menus horizontally
(note that this could be fixed with a bit of delay). He referred himself
to the Ubuntu dock as something better as it's always visible and
accessible. So I showed him the Activities overview: he'd never seen it
and was not convinced it would speed him up (though we didn't test it).
[1]:
https://www.nngroup.com/videos/steering-law/
>>> To balance this, let me make it clear that maintaining the status quo
>>> does not exactly come for free either: we've been paying a recurring
>>> cost for keeping the Applications menu + desktop icons in a good
>>> enough shape. For example, the last iterations (Buster) took us hours
>>> of work.
>>
>> I thought that the Applications menu and desktop icons were orthogonal
>> issues. We don't have the desktop icons because we have the Applications
>> menu.
>
> For me it's not entirely orthogonal: we have desktop launchers because
> that's the only place where we can give them this much visibility,
> which is a constraint that comes from using the Applications menu.
> Other available tools come with different features and constraints.
>
> The Activities overview has this dock with favorites; we could move
> our current desktop launchers there;
The Applications menu has some favorites too and I created #16990 to
update them.
> if needed we could make the dock
> always visible (like Ubuntu does).
I didn't know this was possible, thanks! I mentioned this in #17077.
> And then we would have one single
> place where launchers that we particularly want to highlight live,
> instead of two currently (desktop + favorite apps). Incidentally, this
> could also allow us to trade some vertical screen estate in favor of
> cheaper horizontal space, by dropping the bottom taskbar.
> https://redmine.tails.boum.org/code/issues/11717#note-13 has pointers
> wrt. this idea, for whoever want to explore this further.
Yep. Another issue I have with the Activities overview and this dock is
that their icons come with no label. As a first user, I have to go over
each of them to learn what they mean, while, with the Applications menu
I get the text of the Favorites on first click.
Icons without labels is a well discussed bad habit.
> segfault wrote that #16981 took him "quite a lot of time".
Granted. I forgot about this one.
I'm afraid that with the Activities overview, the discoverability and
usability of anything we have under "Utilities" would be worse than if
we hadn't solved #16981, which was polishing to me.
For example, to discover the System Monitor with the Activities overview
I have to:
- Click on "Activities" in the top-left corner.
- Understand the 9-dots icon in the bottom-left corner and go all the
way there.
- See and understand the gray-on-dark-blue tab in the middle-bottom of
the screen, go there, and click on "All".
- See and understand the vertical carousel dots in the middle-right of
the screen to know that there's more stuff and I can scroll. Carousel
dots are another well discussed discoverability issue.
- Scan 3 screens with a 6 x 4 grid of icons with many names of
applications cut off on the right because they have more than 10
letters, including "System Monitor".
I'm probably exaggerating a bit here but I need to move my cursor
through all corners of the screen, understand and use 3 widgets which
are not super clear, and guess the last letters of the apps.
I'm afraid that what might seems not such much more effort to us but
it's definitely more than the Applications menu and would, for example,
be quite a lot more and a real discoverability issues for people with
poor sigh and slight motricity problems like Cris or my mum.
Don't get me wrong: she knows COBOL but she can't use a computer like
she used to anymore :)
--
sajolida
Tails —
https://tails.boum.org/
UX · Fundraising · Technical Writing
