You might have noticed that we are working on Activities. So far we have done quite some qualitative user research via our blog. Now we think it is the perfect time to move from user feedback to working with the UX experts. So I invite you to join our journey!

Prequel
Large numbers of users do not understand activities. They do not get what activities are for, they gain no benefit, they cannot differentiate them from Virtual Desktops. Also they are kind of broken, which obviously makes learning hard. But at the same time there are fans, that even take the down sides for the benefit they gain.

The underlying problems
(P1) We all are doing different things with our computers. And we all transform our computer into the tool we need for each of these things - sometimes by simply starting a certain program, sometimes by changing some settings and sometimes with complex screen arrangements. And as we users are pretty habitual in what we do, we do the same transformations again and again.

(P2) For some tasks we need more space than our screen offers us.

Personas
These problems are obviously more or less relevant to different users. I guess the more structured and busy users will benefit more form a solution to the problems. But in terms of KDE Personas, I would expect that all Personas will benefit. As it also makes sense to make a more challenging Persona the primary Persona, overall I suggest Berna (structured, busy and cautious). All KDE Personas should be involved as secondary Personas though.

Vision
We will enable Berna to transform her computer into the perfect tool for each of her tasks. She will easily switch the fitting tool as she moves from task to task.

Naming the children
As first step we need to find suitable mental models and fitting names, we can use for communication (internal and external) and streamlining about what we do. Following are some concepts Berna will have to understand, aka have a working mental model of:

(MM1) The upmost idea Berna has to understand: she can switch between different independent
-- states, morphs, transformations, workspaces, activities, Desktops, Computers
-- (all optionally also "virtual")
that all save the current situation.

(MM2) Now Berna has to know that she can change the
-- configuration, settings
of each of these MM1 individually for something she does repetitively - but at the same time some configuration will also affect all MM1

(MM3) Last Berna needs to grasp that as part of MM2 she can create
-- Desktops, screens
-- (all optionally also "virtual")
she can use for complex tasks to virtually enlarge her display.

What do I want from you now?
Feedback and ideas on everything. Esp. interesting though would be to find names for MM1 to MM3 that reflect their mutual dependencies and help Berna to build up good and working mental models.

(Please, I know it is hard, but: no mocks yet. Let's please stay theoretical for a little while!)

bjoernbalazs wrote:but at the same time some configuration will also affect all MM1

This is, in my opinion, the biggest problem (besides being a little broken ATM) with activities as a concept. I have read that it would be very difficult to implement, but if activities are meant to organize and compartmentalize your workflow, and allow you to configure them to suit your specific needs, having some configuration settings be global and some activity-specific is confusing at the very least. IMO they should be like a new session (user), but sharing the home folder and without having to log out/in. Particularly interesting would be activity independent panel arrangement, widgets, power settings and hardware acceleration (an activity without effects for gaming, for example).

BTW, this is related to a thread I started some time ago, talking about ways of showing features that are currently hidden. Specifically about activities, I pointed out that the current interaction with them is not natural: through keyboard shortcut is not very discoverable, and when clicking on the activities icon (if it is on your panel) a hidden panel appears on the left. At the very end of the thread I proposed turning the activity switcher into a plasmoid that could be pinned to the panel, and made a mockup. Here is the link:
viewtopic.php?f=285&t=125710&p=332844&hilit=super+cashew#p332844

I'd like to reiterate what Pedro said.

I only recently started using activities again, am convinced by the concept but not by its implication.

The idea that a new "activity" or whatever you wish to call it starts off with a default desktop as it would for a new user is excellent, methinks. Everything should be totally independent of each other. Alternatively one should be able to copy one's current session as is possible at the mo.

As a concept I don't think this would interest Berna as she is most probably used to Windows and has to manage with one desktop only. But hey, what do I know

Hi Björn,

(MMI1)
I have one desktop at home, one desktop at work, one mobil with me, one tablet for surfing and I have different users on my desktops and the tablet. For me the old activities are the future to convergence (as win10 and ubuntu will have). I'd like to have one device with different activities one for home desktop, some for work desktop, one for mobil and one for reading. Let's say my desktop is a mobil or a mini computer I will take the device with me and for different needs I will have different desktops.

So how should I describe all this different needs with one word. Activities has the problem that it is to abstract for the user. An activity can be everything but what in special? States mean for me that it isn't finished. Morphs is cool but I can't explain my father what a morph is. Transformation is nice, but for software I'm not sure it is the right word. Computers is to complex. Most users are happy when they understand one computer and than they have several.

I like workspace, because it means that you have different tasks and needs for different work the only problem is the name work. It is right that I most of the time work on my computer but people don't associate something positiv with work. I prefere Desktop, because ALL people knows the Desktop in Windows and Linux. But with the name desktop we have a problem with virtual desktops. How to explain the user what's the difference between desktops and virtual desktops. I would say nothing. Merge them together. Than we also have less complexibility. But when we merge them we have to discuss if there are examples where the user need both. I understand that the dev's don't like a merge because this would be a lot of work.

(MM2)
No problem, in the system settings there will be an additional button for apply for current activity and the existing one would be apply for all activities. maybe for the current activity ther button has a drop down list where you can select additional activities for the same setting. When the setting is not available for each different activity than you only have one button apply for all activities.

(MM3)
When you merge activities with virtual desktops the behavior is the same than now for virtual desktops. No new workflow you only have to give the user the information that you can have different settings for the different (old virtual) desktops. So it is more an extension of an existing (well known) solution.
She can use for complex tasks to virtually enlarge her display. As you wrote for complex tasks she can enlarge her display. We shouldn't have activities and virtual desktops only for complex tasks. KDE/Plasmy simple by default, powerfull when needed. So what are the complex tasks where you need both concepts and we will find a solution for this special tasks.

What do I want from you now?
Hope I could help. Mockups later

MM1-3 are easy to understand once the concept has been internalized. However, novices see activities as "meta-virtual desktops that allow you to group desktop configuration and applications together" [1]. The question is how to transform the current knowledge (something like "next to my two screens I can have similar virtual desktops") into a mental concept of activities or workplaces. IMHO that's not a big issue when we provide intuitive access and let the user play with the feature.

Remembering back when i first switched to Linux (Gnome 2, not KDE), at first i thought that having multiple desktops was a brilliant idea to help me keep things organised. Within a few seconds of playing around with them though i was really disappointed that whatever i did to one desktop happened to all the others too, all the widgets, shortcuts etc. etc were always exactly the same. The only thing that they were useful for was putting different windows on different desktops and not 'different setups all within my one computer' like i thought at first. When i started using KDE and had a play with activities it was like 'this is the virtual desktop idea done right' - i could have a different setup on each screen, just not very accessible!

IMHumbleO, Activities should be the primary way to have more than one view of my computer and not Desktops. The UI to switch between Activities could be central to the Present windows function (taking the place of Desktops in the discussion in this thread: viewtopic.php?f=285&t=123481). If you need more than one screen of an Activity then you could add Desktops to each one.

Activities are genuinely useful if you want to compartmentalise your computer and something that no other DE has. One issue is that users coming from other DEs might not realise that changes made to one activity won't affect them all so maybe an 'apply to all activities?' prompt needs to be added when making changes to an activity.

Personally i'm not a fan of the name Activities, especially now Gnome shell uses it for something else but haven't got any suggestions at the moment. Multiple desktops of the same activity could just be called Views - i'd find that more intuitive.

What about renaming Activities as Layouts and making them the primary way of having more than one setup/desktop on your computer, if you need more than one screen for each Layout you can add additional Views of each layout (like Desktops now)?

I'd have thought with a bit of imagination the whole Layouts & Views thing could be accessible through one nice UI.

FWIW, in general I’m with Andreas and Ken on this.

I’m a heavy user of Activities and as they said, it kind of feels like VD’s done right. It might be a risk, to call it that way, but I agree that merging the concepts into one (at least towards the end-user) would be a good idea.

The way VDs and Activities (as they are still called) feel to me is that:

• I see Plasma as my virtual desk(top);
• VDs offer me extra space on my virtual desk(top), so I have more space to work on and where to lay down my tools I need for a certain task;
• when I change Activities, I am changing/cleaning up the desk(top) with all its tools and documents/work on it – e.g. from my work desk to my social one or one of my current big projects

…which is odd and inconsistent with their current naming and just underlines what we’re trying to solve here.

MM1
I agree that we should make Activites the main method of organising your workflow in Plasma. As for its name, I would suggest either just taking over the name (Virtual) Desktop or changing it to something work (in the creative context, not employment) and context related like Workspace or Work environment.

MM2
(I got nothing, sorry)

MM3
IMHO the easiest way to explain this to Berna is that when she is working on more than one project (including private matters and fun), whenever she wishes, she can leave her Virtual Desktop (i.e. Activity) – or even stop it and ignore it until she needs it the next time – and switch to a different project, by opening up a new Virtual Desktop (i.e. Activity) and concentrating on the task at hand, without fear of losing her work or tools (after that actually works better). In each Virtual Desktop (i.e. Activity), she can arrange her tools (applications, widgets, …) and work (documents, data, …) as she likes it – and even add more space for these than what her actual monitor would normally allow for (i.e. create/remove new Virtual Desktops as we call them today).

hook wrote:FWIW, in general I’m with Andreas and Ken on this.

I’m a heavy user of Activities and as they said, it kind of feels like VD’s done right. It might be a risk, to call it that way, but I agree that merging the concepts into one (at least towards the end-user) would be a good idea.

The way VDs and Activities (as they are still called) feel to me is that:

• I see Plasma as my virtual desk(top);
• VDs offer me extra space on my virtual desk(top), so I have more space to work on and where to lay down my tools I need for a certain task;
• when I change Activities, I am changing/cleaning up the desk(top) with all its tools and documents/work on it – e.g. from my work desk to my social one or one of my current big projects

+1, that's a good way of describing VDs & Activites!

I like the idea of replacing the name Activities with Environment, i think Work Environment might be a bit long though....

toad wrote:As a concept I don't think this would interest Berna as she is most probably used to Windows and has to manage with one desktop only. But hey, what do I know

That is a good point, if Berna is indeed coming from the Windows world.

In that case, it might make sense to explain Activities to her with the analogy of separate workplaces or desktops/computers, that she suddenly has (at least virtually/digitally) in her power to organise her workflow as she wishes. Never again suffer from a too small desktop!

I agree, if we can come up with a way of making Activities & Desktops easy to understand, easy to access and intuitive then just like me when i started using linux they'll think 'oh look at this amazing thing that i can do with my computer - it's so easy, i never thought that i could do that'. The hard work is done, Activities & Desktops already exist & are brilliant, we need to tweak how they're presented to and accessed by the user (maybe including changing the name) & change the relationship between them to make them easier to understand & work with.

If possible i'd start by adding the ability to create different numbers of Desktops in each Activity like i think someone has already suggested, as far as i can see the rest is just changing the superficial names, UI etc. etc. etc.

toad wrote:As a concept I don't think this would interest Berna as she is most probably used to Windows and has to manage with one desktop only. But hey, what do I know

I actually think that's possible a reason to make Activities & Desktops easy to understand and easily accessible - she won't realise how great they are until she plays around with it & realises it's advantages for her. If we can make it usable for her then she'll use KDE & think 'I couldn't do that on windows!', if we don't the whole Activities & Desktops idea will probably get copied in a future version of windows or osx and everyone will be saying 'look at this great new feature that ms / apple have come up with, why didn't anyone think of it before?' when we already have thought of it, it just needs making really usable for everyone - i think at least trying to make it accessible for Berna has got to be a good idea.

It's all right what you wrote ken300. I'm also sure that activities are one "unique selling point" that we should bring to the user.

As I wrote before push the features from the activities to the visual desktops. is maybe not the best option, because the users know VD's have there thinking about VD's. I think it is an expanded screen. So put there features like power management would be also difficult to introduce to the users.

If activities are a greate feature user should know, than we should push this feature into the standard kde setting. have an activity plasmoid, ... But first we should define what activities are and what the user can change in each activity and what is a global setting. maybe we can start with what the user can change per activity and than we get the vision to.

TL;DR
• A description of how Berna can discover Activities by showing the impact of settings on Activities.
• Don't change current names, except Activity which should refer to the fact that it is a different instance of the entire workspace.
• Also, keep in mind that converged form factors can complicate things since they might seem like the same kind of functionality to end-users.

This is my suggestion for building Berna's mental model (note that I'm using the current terms when describing this to avoid confusion):

I think Berna should already have firm grasp on the concept of Desktops and configuration of that Desktop, as well as screens and possibly virtual desktops as well. So we shouldn't rename anything in MM3, even virtual desktops. Even if Berna doesn't already know about virtual desktops, this term is already widely in use for something else so it would just be confusing if she were to try searching for more information online (as she should be used to doing, being a very precise person).

At the moment, I believe everyone already implicitly uses at least one Activity, namely the default "Desktop" Activity (if not, I believe they should since it should have no negative side-effects even if they never even become aware of Activities). So Berna will already be using the default Activity, which she could configure for her own needs. I think these should just be the normal settings (presumably from the System Settings) that she can already use. Berna will reach these settings herself as she configures her workspace to her needs. This is something she should already be used to. However, I think these settings should have a minimal indicator of whether they are applied to a specific Activity or to all Activities. I'm thinking of something like a different color for the setting label or an color/greyed-out Activities icon next to it (probably best to leave this to the VDG). When Berna goes to change some settings that are relevant to her, she could notice the indicator and perhaps try to trigger a tooltip or some additional information about what it means (possibly because she has seen the Activities icon in other places throughout the Plasma desktop). So some additional information should be available right away that mentions that this is about Activities. That way Berna can try to look for more information on her own, if she wants to. Ideally, there would also be a hint on where to find information about the Activities, but that may be too much information for a tooltip to an indicator (again, up to the VDG). She could also just recognize the Activities icon in the settings from somewhere else and try to explore that part of the desktop, which would probably lead her to the Activities Switcher. The idea with the indicator is to clarify what exactly will be part of the Activity in a non-intrusive way, but it should also pique Berna's interest about what Activities are. The indicator could be applied to other things that are Activity-specific as well, so it's important that this is as minimal as possible.

As for the actual term to be used for MM1, this could be a Workspace Configuration or Workspace Layout or Computer Configuration. Whichever term is used, I think it's important that it refers to the fact that it's a different instance of the entire workspace. When Berna gets to the Activities Switcher, either through looking for more information (locally or online) or by exploring through the desktop, it should show that she is already using Activities. It should also clearly show which Activities are available, which are running and which are not. It should also be clear that it's possible to clone an Activity and that this does not affect her current Activity, which should appeal to Berna as she doesn't want to lose any of her important work. Since it's clear to Berna that she won't lose any of her work by cloning her current Activity, she may choose to do that as her first attempt at more actively using Activities. It may also be useful to have a small tutorial or Help page be offered at the first time you open the Activities Switcher, or it could just be available from a Help button or something. After Berna has cloned her Activity, she will likely want to make sure her original Activity is still in good shape or she just wants to see if they can really be different from one another. At this point, she should start noticing the different benefits of Activities by simply using them.

An important thing to keep in mind, is the convergence of form factors. This should technically be just another configuration option or just the way in which the workspace is displayed to the user, but from the user's perspective this would be very similar to Activities since it can also completely change the way your workspace or desktop looks. Some may actually want to use a different Activity when they change form factors, because they may want to have a different configuration when they use netbook or plasma media center as workspace. Some may specifically want the same Activity to use across the different form factors and some may want a combination of both. This could get very confusing for end-users, so it might be useful to take this into account already. Perhaps an overview of the exact configuration used in an Activity would be helpful, possibly in a similar way as KInfoCenter. That way Berna, being a precise and detail-oriented person, has an easy way of getting confirmation about everything that the new Activity could change, including settings, active applications, widgets (and their locations), etc.

I would suggest to actually push only one aspect of activities at the moment:
Act when the (hardware) enviroment changes:

User plugs in a mouse, keybord, external monitor on a laptop, ask the user if she wants to use a different enviroment for workstation setup.

User plugs in a second monitor, ask the user if she wants to use a different enviroment for dual monitor setup

User plugs in a tv, asker her if she wants to use PMC

...
maybe give the option to switch activity on a specific wlan?

And the next time the same hardware is pluged, auto switch to the activitiy. Ubuntu, MS start doing something similar, so that will be a known concept. It's triggered by a comprehensible event.

My personal opinion regarding different activities for different tasks/projects, don't push this feature at all: Its only benefitting for some (few?) workflows, it's a concept that's very alien to users of different desktop enviroments, there is no clear boundary to vds.
Keep it a s a feature for power user, but don't promote it.