{December 18, 2010}   Activities 4.6 screencast

Yay! Screencast! :)
For 4.6 I decided to do two separate screencasts: one intro to the basic features, and a separate one for cool advanced stuff. This is the intro, naturally; I’ll do the advanced one tomorrow or something. :)

…There’s another thing different here: I’ve enabled ads. I’m not sure about this – on the one hand ads are annoying, on the other hand a decent screencast takes hours to make and a bit of funding for my kde stuff would be useful. The ads haven’t started showing up yet anyways. (I guess it takes a while?) So, if you do see ads, tell me what you think of them.

Now on to the screencast:
[Edit: ads are off now]

Tom says:

Whats up with all the friggin ads before KDE videos??
Blip sux.

teho says:

You could have read the post first.

So now that activities support all(?) the features that the virtual desktops do, what will be the future of those in the KDE?

Chani says:

not quite all: there’s no shiny composite effects for switching activities. ;)

I still use virtual desktops to have more space within an activity. Usually the first desktop is for work, the second is for testing, and the third is for distracting cross-activity apps that I need (mail, irc, etc – I really hope to have activity-aware clients by 4.7 so that they’re less distracting)

I expect some people will stop using virtual desktops, but I also know plenty of people like me who need both. :)

TZ says:

Add direct link, please:

[video src="" /]

Yuri says:

OMG, ads of new windows phone! :(

BTW, thanks for the screencast! ;)

Tsiolkovsky says:

Yikes! I got some propaganda from evil proprietary corporation and their new Vista 7 Phoney. I think it’s better to turn off these ads if this is what we get.

Martin says:

Chani what window decoration/border are you using??

Just curious!

Chani says:

just oxygen. :) I might have fiddled with the titlebar settings a little, I forget how…

Martin says:

Thanks :)

Jeffery MacEachern says:

The relevant option is Configure Decoration->Fine Tuning->Outline Active Window Title.

Jeffery MacEachern says:

Er, wrong person, sorry. But the comment still applies. :P

xian says:

First off, thanks for all your work Chani!

I use a linux desktop at work. I currently use multiple (four) desktops. One for terminals and a web browser, one for my windows VMs, one for my current projects including text editors, web browsers, and terminals, and the last is miscellaneous. I have things like my mail and im clients show up on all desktops.

Currently this seems to work well and I can use keyboard shortcuts to jump between desktops or use the mouse scroll wheel.

Based on your screencast I’m not exactly sure where activities fit in. I suppose for things like misc. work, I’d be able to run as a separate activity and when that job is completed, kill off the activity and all apps associated. Any suggestions? How do you use activities? I could see it possibly being more useful on a laptop for home/work activities; pause one, travel to the other environment, and activate the appropriate activity.

Jeffery MacEachern says:

“Based on your screencast I’m not exactly sure where activities fit in.”
xian: What is shown here is one of the primary user-visible manifestations of Activities right now, and that is probably why people keep confusing them with virtual desktops. However, they’re a fundamentally more powerful concept because they allow any application to take actions that are useful to react to your current “activity” in a general sense, rather than simply organizing things spatially.
As an example (one which I didn’t manage to get into 4.6), consider the “Places” sidebar in Dolphin, etc. As a university student, I find it handy to keep shortcuts to my folders for coursework in the sidebar. However, it starts getting cluttered really fast. Activities could allow you to specify that your coursework shortcuts should only be shown when you’re in the “School” activity.
It’s up to application developers to integrate Activity awareness in useful ways, so if you can think of somewhere it would be useful, go talk to the developers behind your favorite app. :)
In brief, what we have now is essentially “Plasma and KWin integration” with Activities, but it’s not limited to these uses.

Chani says:

exactly :) :)

Chani says:

hrm. I was hoping the screencast would explain it all… you have the right idea, putting different jobs on different activities. It’d probably be most useful for splitting up your “projects” desktop. I also have one “misc” activity for things not big enough to be worth their own (eg. looking up some restaurants; not something I want mixed up with work websites but not something I’ll keep around long either).

I have my mail client on all activities, but never liked it on all desktops for several reasons: first, because sometimes I have programs that work better with a desktop to themselves (eg. the gimp – lots of little windows that you want to see all at one) or I need the desktop exposed for testing plasma stuff; second, because it’s easy to be distracted by offtopic mail (sometimes I take it off my work activity; I’ve already confined akregator to “fun” even though it has plasma-related things); …and there was a third reason but I’ve forgotten it. :)
*yawns* I think it’s bedtime. :) I’ll try to show more of how I use activities in my next screencast. Also, I think RC1 is coming out on wednesday, so you could play with it yourself then :)

Cornelius says:

The sreencast is nice, the ads are horrible.

dodosoft says:

Great work!

I agree that Virtual Desktops are still useful to distribute windows but unfortunately the VD concept has been expanded and now you can have different widgets on different VDs (plus having different background images etc…) so now VDs are nothing but “another kind of activities” which is totally nonsense.

Now that Activities are taking shape, please remove that feature for VDs and make everything more rational.

And PLEASE, make panels part of an Activity, let people set up different panel configurations for different Activities.

Chani says:

haha, a lot of us want to remove that “widget groups per desktop” feature, but there are a few people who got attached to it…

there are still plenty of other features that vd’s don’t have, though. and like I said, they’re useful for having more space within an activity.

yagami says:

Yeah and i am one of them.

currently use PVD on my kde desktops.

why is the need to remove VD’s features to explain why activities are good ?

and PVD isnt even a VD feature. i think of it as a specific focus within an activity.

Media activity works best with this, as there is a PVD for music and another PVD for video.

i guess some people only see what they have in front of them and cant be creative about the possibility of activities.

i dont even think of VD’s in PVD as virtual desktops. i think of it as diferent works within one activity

javi says:

Currently the only useful use that I see for Activities is to have backups of my customized Virtual Desktops so if by precaution I remove something or I me strange experiments with widgets I can restore it clonning the backup activity and removing the old.

It only needs a “independent panel per activity” option and a clone of the existing Virtual Desktop manager widget that should show only “enabled” activities and then I could use more the activities for other uses but for now it is only a backup tool.

Chani says:

as for panels: patches welcome. :) it’ll be a useful feature, but it’s not trivial to implement (if you want it to be reliable), and nobody’s volunteered to do it yet.

Rsh says:

On the screencast plasma looks really sluggish, it might seem a showstopper for people starting their KDE adventure from screncast. :P

Chani says:

well, that’s always a problem with screencasts.

it’s not usually *this* slow, though. hrm. my poor computer needs some cleanup, I guess…

foo says:


I don’t want to sound negative, but I think this is a lot of work (plenty of clicks and typing) just for opening and closing some windows. Seems like named sessions or more inconvenient virtual desktops and nothing much.

Yes, I know, you have plans for the future with nepomuk and stuff. But if I were you, until you have something that really makes sense, I would stop making this screencasts.

Sorry, I know that this seems terribly negative, and don’t get me wrong, I usually love everything that comes from KDE, and probably in 4.8 or something I will be the happiest activities user on the planet (I hope I will), but untill that happens, I completely fail to see how this is useful. :-(

Chani says:

er… yes, you do sound very negative, with no constructive suggestions (don’t worry guys, I won’t stop making screencasts), and I think you missed a few points too. like the part where I closed the whole activity and opened it up again. you can’t do *that* with virtual desktops.

as for the number of clicks… that doesn’t apply to opening and closing windows (I could’ve put a konq launcher on my panel instead of using krunner, that’s just personal preference) but there are other parts of the activity workflow (eg. moving a window from one activity to another) that I’d like to improve. Normally I use the activity-bar plasmoid, too. That’s something I’ll show off in the second screencast… :)

Markus says:

Actually, I don’t find, what foo says, negative, I think he is hones, sincere and has a point, actually I think exactly the same.

I really love what you and the rest of KDE are doing, and I can imagine that activities might become very useful, but right now all this looks terribly complicated, whereas it is not clear, what should outweigh all this.

Maybe you should clearly communicated, that the usefulness is something, that will come in later versions and that this is simply a neccessary first step.

Keep up the good work!

Wojtek says:

hi Chani,

I must admit you have nice voice and I think it fits your name :D

Jeffery MacEachern says:

Chani: I didn’t see the ads, since Flash decided to break again recently, but you could also consider trying out the Flattr service, as a lot of other FOSS people have. It’s quite nice to use, although there is of course the catch that you have to reciprocate.

I don’t like ads, so I didn’t watch it.

Jeffery MacEachern says:

See the OGG video link from TZ (above) – it doesn’t have ads

crabman says:

For changing the activity one has to move the mouse to the top right corner, click on the cashew, click on “activities” then move the mouse to the bottom of the screen and click on the desired activity.
Way too many clicks – if anyone manages to find this feature he will never use it because its inconvenient.

Sorry for the harsh critique but thats what i think everytime i read about the activity manager. To make up for that i have a suggestion:

when the cashew is moved to a screen edge it looks like a tab and shows the activity name. Why not use it as a tab? add a second tab with a small plus sign to add a second activity. Once its added, show a third tab for the new activity and so on. The tabs could have context menues to rename, change icon, pause …

This way the activities feature is much more visible, faster to use and the cashew finally has a raison d’etre.


Chani says:

ahh, *that*. :)
yes, I don’t use the activity manager for switching. I use the activitybar plasmoid, on an autohide panel at the top of my screen. :) maybe I should’ve put that in hte basic screencast after all.

there’s also a mouse plugin for changing activities, and a keyboard shortcut (meta-tab), although I don’t use those much.

oh, and we discussed the issues with the activitymanager UI a little while ago, and there might be a better one done in qml for 4.7… but in hte meantime there’s the activities dataengine, so anyone can write a plasmoid for activity stuff. :)

KDE user says:

I would love it if it were possible to change between activities with a cube like effect (just as it is done with the virtual desktops atm).
When an activity is paused a gray-scale screenshot of prior use of the activity should be displayed instead of the actual contents of the activity and when changing to it it should be started automatically.
Load-based stopping of hardly used activities would also be awesome.

Xehoz says:

Activities seem to be progressing well. I sort of stopped using them though. Making use of the virtual desktops with different plasma elements is usually flexible enough. However, the screencast for 4.6 shows that activities are stepping up a bit.

«For changing the activity one has to move the mouse to the top right corner, click on the cashew, click on “activities” then move the mouse to the bottom of the screen and click on the desired activity.
Way too many clicks – if anyone manages to find this feature he will never use it because its inconvenient.»

There is actually a quick way to change activities in 4.5 (create a blank small panel and add the activity switcher there). However, you can’t activate and deactivate activities like Chani did in the screencast, unless that plasmoid also got an upgrade. But I tend to agree with what you said: switching and (de)activating activities should be straightforward.

There’s a recently released plasmoid ( that can do the trick.

Chani says:

ooh, I didn’t know about that :)
sadly kde-look seems to be down at the moment. I’ll check it out later.

Blackpaw says:

Still not really seeing the point of activities – very little that can’t be done *easier* with virtual screens.

For me, until I can assign a hot key to each individual activity they will be too awkward for switching between. The plasmoid is useless for that – isn’t usable vertical and takes up to much screen space.

“It’s up to application developers to integrate Activity awareness in useful ways”

Kinda of like nepomuk, which to date has very little visible use to the user.

Jeffery MacEachern says:

“For me, until I can assign a hot key to each individual activity they will be too awkward for switching between.”
Meta+Tab already will page through them, although I’ll grant that it’s less than ideal.

“Kinda of like nepomuk, which to date has very little visible use to the user.”
[Disclaimer: I haven’t looked into Nepomuk’s API to any extent], but the entry barrier to using the Activity API as it stands is trivially low. IIRC the API hadn’t stabilized until recently, so there wasn’t much of a chance to make use of it. I expect usage will pick up; I know I’ll be looking for ways to make use of it. In the simple case, it can be as easy as a couple of method calls and hooking up a Qt Signal to listen for Activity changes.

arwa says:

I do have this problem, too. I cannot see the advantage of these activities over the virtual desktops. It starts with the name that is hard to understand. I am not sure what an “activity” should mean in the desktop context.

Also KDE is adding another dimension to virtual desktops with the activities. It is very hard to sort out when I should use a desktop and what for an activity. How should I combine these and when to switch what?

I am very interested in KDE things and always try them and want to understand them to see if it might be a good idea for me to use these things. But activities are too complicated IMHO and I do not see the advantage yet.

Why not offer an easy way to define different wallpapers for different desktops (like nearly every other WM and DE) and to bind windows to not only one or all, but to some virtual desktops without the need of additional dimensions of activities?

Well, I will not be able to stop this confusing movement to activities, so I will watch the future development and see if it wil become usefull for me one day.


Blackpaw says:


Artificial issues as far as I’m concerned. Spatial layout? I never even consider my desktops as having a spatial layout – they’re a virtual workspace I switch to – usually via a keyboard shortcut, that’s all.

arwa says:

Thank you, chani, for the link. I will write a comment there. (Why isn’t there a reply link in your comment here? Ok, I now reply to my posting.)

arwa says:

Oh, comments are closed on the linked entry. So I comment them here:

“1) you can only remove the *last* desktop.”
…perhaps you could hack around this by shuffling all windows from higher desktops over one and renaming all those desktops and then removing the last one. it’d be icky, though, maybe slow, and probably prone to glitches.”

Well, is the handling of activities so much different for the system? When switching, removing, reconfiguring an activity the windows probably must be rearraned, too.

“2) adding a desktop messes up their spatial layout.”

This is because the adding feature is conceptional not correct IMHO. The system should offer a way to add one row or one column to the desktop grid. This point is not a disadvantage of virtual desktops, but only shows that the grid idea is not complete yet.

“3) a window is either on one desktop, or all of them.”

I am sure that this could be implemented easily (not by me, as I am not a C++/KDE coder). There is no difference to activities that had to get this feature implemented, too, before it did exist.

“4) compositing exposes lies.
imagine that kmail has the ability to only show mail folders related to your current activity.”

Will this be possible? It might get very confusing. But ok…

“now imagine the desktop cube showing kmail on two desktops at once. kmail doesn’t know you’re using a compositing effect; it’ll display as if it were in the current activity, both times.”

Will this change by activities somehow? This sounds more like a shortcome of the activity concept than of the virtual desktop concept.

So I think that these mentioned drawbacks for virtual desktops could be solved cleanly without dirty hacks. The result would be very clear for the end user to understand. Adding activities does not fix these things, but only offers more complexity. Don’t they?

Chani says:

““3) a window is either on one desktop, or all of them.”
I am sure that this could be implemented easily (not by me, as I am not a C++/KDE coder). There is no difference to activities that had to get this feature implemented, too, before it did exist.”

uh.. dude. I *am* a c++/KDE coder, so trust me: it could not be implemented easily. it’s kinda arrogant to assert that something must be easy while admitting you don’t know how it works.

not that it’s impossible, but it would be a pain, and break lots of applications that try to be smart… actually we’re seeing some of the shortcomings with activities anyways, but at least this way we don’t lose the desktop stuff as well.

arwa says:

Hi Chani,

“uh.. dude. I *am* a c++/KDE coder, so trust me: it could not be implemented easily. it’s kinda arrogant to assert that something must be easy while admitting you don’t know how it works.”

I do not want to be arrogant. Sorry if I sound like this. (If I wouldbe arrogant, I would not mention that I do not know coding KDE/Qt/X.) I am a developer (in the area of web and perl) so I think I have a feeling of programmatical logic and what might be possible.

In this case of a window on multiple desktops it sounds very unlogically to me if it is possible to have a window on multiple activities that can be assigned to multiple desktops, but not to be able to have a window on multiple desktops directly.

If you implement an hidden activity layer and relate activities and desktops in a 1:1 manner, you could exactly do what you are doing with activities now: Relate a window on different desktops.

But ok, it is a black box for me. Maybe I am wrong here.


Chani says:

:) but you’d have to lie to the applications and let them think they’re on desktop N, which would have funny side-effects. plus I wouldn’t count that as ‘easy’ anyways…. the kwin source code is a bit.. daunting at times ;) although I guess that makes activities alone not easy either.

anyways, my general conclusion was (and still is) that while it’d be mostly-possible to jam the two concepts together, it’d create lots of weird annoying side-effects, and besides, it would prevent me and several people I know from working the way we want: lots of activities with a few virtual deskops.

you can still have only one virtual desktop if you like, or only one activity.

molecule-eye says:

I feel you. I can’t see certain advantages of having activities but I can’t see that, overall, I should switch to using activities from using virtual desktops (for certain purposes). For one, it’s way easier to switch VDs and to move various windows around from one VD to another.

Am I right in thinking that activities are trying to simulate the effect of having multiple users accounts mashed together?

Chani says:

activity switching and assignment of windows have room for improvement, yes (although there are already far better ways of switching than what’s shown in that screencast).

as for multiple user accounts… while it does bring some of the benefits of multiple accounts (like sessions) it’s not meant to simulate that. It’s more intended to bring the best of both worlds – so you can separate your projects while still having common things (eg. mail client, bookmarks) available everywhere.

brian mca says:

For me, until I can assign a hot key to each individual activity they will be too awkward for switching between. The plasmoid is useless for that – isn’t usable vertical and takes up to much screen space.

“It’s up to application developers to integrate Activity awareness in useful ways”

Chani says:

hrm. I feel I should explain another thing – on my computer, I don’t want switching activities to be too easy. I use activities to keep distractions away, and if I was constantly bopping between ‘work’ and ‘fun’ it would defeat the point. ;)

I don’t want them to be too inconvenient either, so I use the activitybar plasmoid instead of the manager UI, but I find that the little bit of extra effort involved in using the mouse is enough that I don’t just wander over there unconsciously.
…or maybe that’s just because I’m not used to it; maybe I’ll have to remove the activitybar later because it’ll have become too easy ;)

…anyways, I’m pretty sure it would be possible to write a plasmoid that lets you set up global activity shortcuts.

Fri13 says:

I as well do not see any points for activities. Not after this screencast or the blog of the Aaron Seigo who wrote about differences of GNOME Activities and Plasma activities.

I must say that I like very much more about the GNOME way as it follows the very smart and logical virtual desktop ideology.

Like since 90’s I have used virtual desktops to separate the tasks what I do. I could have added or removed the virtual desktops to meet my needs.

Virtual desktop 1 for the task 1
Virtual desktop 2 for the task 2
Virtual desktop 3 for the task 3

Now the current activities very much clash with the virtual desktops idea totally.

If user wants to do something, first thing he/she needs to do is to create a new activity or open the correct activity.

I am a human, as any other computer user.
I do not want complex things, as almost all basic users. Even that I am advanced user, I want simple things and I want that basic users gets computer as a tool what helps them to achieve their goals, not to tweak the tool because it is a tool.

I am a multitasking person. As are most other people. I need to do small quick things right now. Like if I am writing a work document having a browser behind the word processor with few tabs and someone asks can I check what is the next movie what comes in the theater and when it starts.

I do not switch to other virtual desktop or definitely I am not going to switch activity to “fun” or “home”. As I can just simple press Alt+Tab, Ctrl+T and type the movie theatre address or pick it from bookmarks and check the info and close the tab with Ctrl+W.

It is much faster to do just the Alt+Tab now than even using a Meta+Tab to switch activity what would be needed to get the whole thing working.
I already have a browser open behind so why I should switch even virtual desktop or any activity to do fast and simple thing?

There are some points what I can find activities really important and made from gold.

But they are only for very very specific needs. Like as someone said, a student who needs to get access to only needed shortcuts in dolphin.

And currently I already have that but not because activities, but because of virtual desktops.

We already have great widgets for desktop like folder view. Just add needed folder views including the needed shortcuts in it and you have a own virtual desktop assigned for specific work/class.
Right now I have two virtual desktops assigned for two different school project. With virtual desktops I can gather all the needed tools and data to one screen. Separated from any other virtual desktops. When I switch the class, I can simply switch the virtual desktop. But never I am forced to switch activities to get access to other data as it is easy as just switching virtual desktop or using the dolphin to get access to wanted files. With activities I should first change activity and then change virtual desktop and then get access to files.

I bet that most people do lots of small multitasking that activities how plasma will have them does not work at all.

The GNOME idea for activity = virtual desktop is limited. But it really is very great for basic and advanced users. Power users will find it limited but…
Even they have the same basic problems as the plasma activity what is that you have browser open for specific task and someone asks to check something. Or user just simple remembers something important and does it right away. WHy to switch virtual desktop or start a new activity when you have a browser open and you can just open a new tab?

Right now the plasma activity can bring more problems than what it might solve.

It is like user would have 5 exactly same hammers in their belt. And they would need to switch the hammer when they hit different size nail or when they hit the nail for different part of the wood.

It is very clear that such work would come very slow and many would get so angry or lazy that they simple would not like to switch the hammer every time they hit the nail as they do not see the point when all the five hammers are exactly the same! Why to switch hammer just for switching?

I were a heavy virtual desktop user at KDE 3.5 times. Even at KDE SC 4.1 times I used them a lot. But today I have the default setup of the 4 virtual desktops and I actually use just one.

I have development applications, graphic applications, browsers, file managers, konsoles and media applications at same desktop.

Why so? There came few very important features to KWin what actually made virtual desktops less needed. And it was what Apple invented for cheetah (?), the exposé function.
I do not need to search window from taskpanel or go trough the alt+tab. I simply wave the mouse to corner and I can see all. When I have computer longer times running and I achieve the 20+ window amount and I can not find the wanted application in two seconds. I simply type the name and it is shown.

Sometimes I want just to focus to the specific task. And this is the case where the Plasma activities should come in handy. But I find it much smarter and simpler to use a the three empty virtual desktops for that task.

And specially now in KDE SC series times (not possible at KDE 3.5.x series) is again the great KWin function what not even Apple did invent. The desktop grid what came from Compiz-Fusion. Just swing the mouse again to other screen corner and you have all desktops shown as grid in front of you.

And KDE developers made very innovative improvement in the desktop grid. That every window open in different virtual desktops, are shown as exposé effect. That you really see every single window open in every virtual desktop. THAT IS JUST SO DAMN AWESOME!

That is such a function what you show to ANY Windows or Mac OS X user and they are sold…. They want exactly that.

And now newer KDE SC releases (was it 4.5) includeds the +/- signs at bottom right corner to add/remove virtual desktops. That is as well very awesome!

And here comes now the point why the Plasma activities are not needed by many users:

1. People do multitasking. They open different web sites, different documents and keeps pauses with different tasks. Listen music or even watch videos at same time they write something, they have IM chats open, IRC open, emails from different people.

2. When user wants to focus for some task, they simply open desktop grid and they move wanted windows to that virtual desktop and then switch to it. Then they can fast way switch virtual desktops if wanted.

3. If (when?) the default 4 virtual desktops gets filled up, user can just open again the desktop grid and add few more.

All the time the usability is very simple. The desktop grid is the “activity manager”.
The users can mix different tasks all the time. They do not need to remember on what activity they were and be worried that they can not open the website, answer to IM, write a email or a add something to existing document if it does not belong to that activity.

The plasma activity and virtual desktops clash to each others and I see that there comes more complexity to the user for desktop management than what they can gain from it.

Activities would work for some people, with narrowed working styles. Like when they start to write email for co-worker or report for boss. They close the music player, they close the web browser and all non-needed applications from that task. And these tasks would be repeated all the times. So it is nothing like “write a document once a week/month” but every day or so. So it would be worth to actually create and customize the activity what to use.

Not even activities based to locations is so nice. Like activity to home and activity to work. Many people do work at home or home things at work. In paper everything sounds very nice. But people do check their work emails at home, write some documents or finish the presentation. People chats and writes to family and friends while working.

The plasma activities would work better if they would be tied to virtual desktops by default. That every virtual desktop is different activity. But even then user has multiple tasks on same desktop (on same activity) at same time.

User does not want to open or switch first to correct activity and then open the application for that. Like if user gets email or IM from friend, it is silly that to read it you need to switch activity. Or that you could not receive them if you are on other activity.

I wish that every plasma developer would attend to activity brainstorming and discussion. That how it really would work at different basic situations (IM, Chats, multimedia, web browser at different task etc etc etc). And then actually focus to bring very accurate basic user situations where the activity would help users and where it would bring lots of problems.

As I can see more problems or complexity with activities than fixes and help.

The question remains, why I should start using a activity manager when I can very well now do everything just with snap’n’drag way with virtual desktops? Even session control is handled and KWin allows to apps open specific virtual desktops etc etc.

(The GNOME idea that actually in the end as well force to have one app open per virtual desktop is as well very terrible. We need a way to mix every task with every virtual desktop or work. People do that and they will continue to do that no matter of what).

damipereira says:

Your post gave me a new idea, what about making a desktop grid where every file of desktops is a different activiry, when there’s no extra activity activity manager behaves like desktop grid, providing buttons right below the first line of desktops to add a new activity, or turn on powered off activities(1 button per activity)
This activity icons, when clicked would expand into another line of virtual desktops, and behave exactly the same way that activities do now.
After reading a bit, I feel like activities are a groups of virtual desktops made to be specialized at one task.
It isn’t neccesary to change activities when doing something trivial, but it would be nice to have an activity for doing specialized things, like work/programming/fun
With this new desktop grid it would be possible to easily move windows between activities, to change to a specific desktop on the other activitiy and the most important it would feel cleaner than actual workflow, I’m not saying you didn’t do a really good job but I think the interface could be improved.

Chani says:

hrm.. I have a vague memory of considering something similar (VDs on one axis, activities on the other). I can’t remember why I dismissed it, though.

…of course, it’s entirely possible for someone to write such a plasmoid and try this idea out. :)

T_U says:

I didn’t follow everything, but some say they don’t see the point because they are basically multitasking humain beings, doing everything at the same time.

THIS IS EXACTLY THE POINT in my opinion. Too frequently, I try to do everything at the same time, which prevents me from focusing on a single activity and doing my work properly. It’s too easy to alt-tab or switch workspaces.

So, I don’t know, it may turn out be very interesting…

BTW, I’m kind of dubious about GNOME activities. I saw a video (this summer) showing that activities would be (tell me if I’m wrong) a kind of virtual desktop showing a single application that has several windows and a global menu for all those windows. I’m really dubious about it. One activity is almost never made of a single app. Dunno.

The good point is : it makes us all think about it :)

Chani says:

+1, we think we can multitask but we can’t :)

as for gnome, I went and did a it of research. they do not *have* activities. what they have is an overview screen that they wanted to call “activities overview” but there was only space for one word, and they picked the first one. :)

Fri13 says:

Yes, we can not do real multitasking (but neither could desktop computers if person did not have SMP (dual CPU system) or CPU with multiple cores) but as the original multitaskin from computers is just that the operations are switched very fast rate, so the people believe computers runs everything at same time.

Humans do exactly the same multitasking, they switch operations.

And I did not mean that I belongs to people who believe they can do multitasking (as I know it is better to focus just for one task at time), like over 70% people believes but in reality only under 2% of people can do it and gaining work speed with it, while others loose time.

But what I meant, was that we really need to do multitasking everyday. And we really do it still pretty well.

Example. We drive a car and we listen music or we talk with passanger. We can do it easily as we can stop listening radio when ever wanted and focus to driving. And the passanger focus as well their talks to the drive situations (most of them know to shut up in important positions who knows how to drive).

But when we are talking to phones, the other person in the phone does not see what situation driver is. And the driver needs to focus to listening the voice and holding the phone. (Headsets does not drop accidents amount).
The problem is that the person next to us can change their voice volume and talk speed acording the situation. And that the passanger talk does not disturb so much (psychological proof for that).

But we really do multitasking on those times. As we do with accelerating/breaking and turning the car. No, not all people are good at that either, but most can do it. But when we get third thing to situation, like accident why we need to do fast choice and break. We loose 1 second just for that reaction to understand what is happening.

With computers desktop. We really need to do a full time multitasking (even that there ain’t real multitasking. Only marginal group of people (1000 or so in world can look two different directions. 10 000 can write two different text with both hands at same time but only from memory. Not by listening two person and write them to paper) can manage on that truly.

But in daily purposes. We are less alone in one room writing something. Most times we use computers at phone, at office with shared offices (or cubes), home with roommates, TV or Radio on and so on.

There really ain’t many situations where we focus just for one task at the time. Activities work for those people and they are marginal people, at marginal situations.

Is there any activity developer who can say that they focus just for one task at time? Without need to change the activity until they have completed the task what they are working?


If I write a school document or report to boss. I usually get lots of phone calls or some asks from next the room/office something. And I need to answer to that. Example look the bus schedule or the movie theather schedule. I can not say “Sorry, I am writing now this answer to the forum and I need to finish it first”. If it just takes from me under 10 seconds to open a new tab, type a site what pop-ups from bookmarks and type the wanted searcha and say what are results, why I would not do it?

It does not actually change the time what I have for the task what I were doing. Yes, it could cause me to loose focus. But then it is my problem if I did not close the door and place the phone in silent mode etc.

The same problem exist in pure computer situations (counting environment away). Like I am writing a school document for next day as I have evening plans with friends and we should meet at our usual place at 8pm. I watch time and it is 3pm so I know I have 2 hours left and I need to horry.

Why I would not like to get a IM or Email from my friend who says he is sick or got other important things to do?
At that point, even that I could chck those before rushing out of the door. It is more important for me that I can just relax for writing the document by spending whole evening for that.

The Nepomuk or Activities could not help me at those points if I would not have that meeting in my calender but it would be just usual known thing.

At that point, I really would like to get the IM/Email and loose the few seconds focus of the document writing.

As for one task, we use multiple tools. Like when writing a document, we use browser, office suite, music player, IM, Email, notes, photomanager, photo editor and so on.

Are there anyone who does just one task at time? Like they do not listen music when they code or they do not sit on IRC when they write emails?

We already have a virtual desktops what gives users the power to focus just for one task at the time.

Place all the tools needed for that task to same virtual desktop and problem solved.
It is not so hard to add wanted folder view widgets to desktop (or panel if the panels could have different configurations at every virtual desktops!) to get quick access to wanted/needed files.

And even then, no need to loose time to solve the activity and virtual desktop mixup or start a new activity just for that someone asks from next room when the wanted movie starts from TV while writing a document.

I do not like the idea what GNOME does for activities. But same time I can not understand why we need to do easy tasks so complex with activities for KDE Plasma as well?

Right now, I have three tasks (or activities as KDE wants to call them) running at same time. I have them separated to three different virtual desktops. The fourth virtual desktop is for global tools, like IM, Music, Videos.

On one virtual desktop right now I have a document processing.
Second has different forums and blogs when I have time, like this one.
Third one I have photo management going where digiKam just finished importing 8Gb of photos.

I do not need activities to manage my tasks. As I can manage my tasks just fine. People who learns the virtual desktops fins just perfectly fine to solve that problem. As they do not separate one task to multiple virtual desktops.

Even at this point, I have multiple browser window open at every desktop and so on.
When I am writing a document, I might get a idea for some other tasks. So I fastly switch to that desktop and I do it until I loose it. Then I continue the earlier task.

If I need to start play around with activities+virtual desktops. I (and I can think most as well) loose the power to actually do things.

So it comes to situations were I need to have four phones and every one for one task. One for calling, one for notes, one for SMS and fourth for address book. When ever I need to do something, I need to search the wanted device.
It is like I could not be on phone and write a note for notebook or piece of paper at same time.

I can not find single work fir what I would like to focus all my time right now.
And the very rare times when I do. I close the music player. I go offline from IM. I do not open IRC or I do not open email client. I just wont simply open any other task when doing something very important.

The activities could work on rare situations. Like we would have a work activity and home activity. And the activity would be network location tied. When we are home, we can not access to home activity. And when we are work, vice versa.

But who really would like that? Does not anyone work at all some personal tasks at work time? Like if you need to contact to your phone company what has call time 09-16 and your work time is 08-16, you do not ever call there as it would be at work time?

We do a lot personal things at work time. As it is allowed. The worker gets paid from the time when she/he is available for the payer.

Many does work jobs at home, without getting paid from it.

Kids does homework at home, not just in school.

People use personal time for work/school and vice versa.
There ain’t any black and white usage for different activities.

The activity developers needs to show us what daily problems the activities solve. I have watched every activity presentation and I can not see more than new problem solvers without problems existing.
What activities really do on daily purposes what virtual desktops does not?
I believe people would not want to see technical oriented information but more the normal problem solvers.

Chani says:

“Is there any activity developer who can say that they focus just for one task at time? Without need to change the activity until they have completed the task what they are working?”

of course not! god, you’re talking as if I’m going to come into your house and force you to *only* think about things related to the activity you’re on, and never leave until you finish!

I switch back and forth all the time; I have the activitybar set up so I just slam my mouse to the top of the screen and click the one I want. I open offtopic windows nearly every day (partly because I can’t filter mail or irc by activity yet), and if they’re for something quick I close them before they ever reach the activity they ought to have been on.

activties are a tool. I use them in ways that are useful, not in ways that don’t make sense. you can be as strict or as lax about their usage as you want.

Jeffery MacEachern says:

@Chani: Yeah, it’s a pity the Activity Police didn’t make the feature freeze. Ah well, I guess we need to step up recruitment and propaganda. ;)

On a more serious note, though, I agree entirely. Users should feel free to use, abuse, or ignore any feature as they see fit. I would also point out that Activities are a conceptual tool at a fairly high level of abstraction. They can be used to /separate/ different tasks (both in terms of direct organization, and cognitive focus), but there is no enforcement of that use-case. We’re not going to *force* users to create a new activity every time they need to open KCalc in the middle of listening to music!

T_U says:

BTW, on the Amiga, one could assign apps to some specific virtual desktop. (and each desktop could have specific settings, even specific settings for each widget etc). Is it possible in KDE 4.X ? It was in KDE 3.X if I remember correctly.

They weren’t virtual desktops, IIRC there were “screens” that could even have different resolution (at least up to Kickstart 3.0, I didn’t go further than that).

T_U says:

Indeed, but there were “public screens” that could be used as virtual desktops.

One could “promote” one app to a public screen and each public could have a different resolution / depth etc. When using MUI (Magic User Interface) one could customize for each app the settings. (promote to a public screens / settings for each kind of UI elements, it was really crazy, for instance, it was like having a different gtkrc for each app)

Also the Amiga could display several screens that had different resolutions at the same time. (that was possible due to specific hardware tricks)

FiNeX says:

I want to share my experience with VD and Activities.

Currently I’m using 6 Virtual Desktops and 2 Activities.
I really like this configuration because I can split tasks on Virtual Desktops and I keep different widgets on the activities. For example VD 1 to 3 are for develpment and browsing, VD 4 is dedicated to PIM and chat, the fifth one for virtualization and the last one for amarok.

One activity is clean (without widgets and any visual noise), so I can focus on windows, The other activity contains _a lot_ of notes and some other widgets. Keeping two activities allow me to be not distracted.

In the future, when windows will be easily associated to specific activities, probably I’ll use activities to better separate group of tasks.

Yes says:

You have a very beautiful voice. It is very pleasant to listen to you. Great screencast!

KDE 4.6 seems to be great!

chromatic says:

Thanks for the screencast !

At the opposite of some users, I really see the point in activities over VD. I’m agree to say that there’s not so much differences or advantages actually.

But like chani said : “It’s up to application developers to integrate Activity awareness in useful ways”. So with a bit of imagination and a mind able to go beyond a 20 year model in computer usage, here some use case :

– Like said before a student that use an activity for each coursework with related plasmoid on desktop
– I arrive to work, with geolocation my laptop switch to the “work” activity and I’m only notified of mail and IM from the people I work with. VD are still usefull for managing a lot of open documents/apps
– I launch a movie from dolphin, nepomuk knows this file is a movie ressource because I told him with bangarang(media player using nepomuk), the media player is launched in the entertainement activity and all types of notification are disables.

This are really simple examples, and the possibility are endless if you don’t have a narrow mind. And if you want to use your desktop in the old way, just use one activity.

I believe it’s a lot of work on the code side but it will come. So thanks to all KDE people for taking desktop to a new paradigm !

Psychotron says:

I’m really looking forward to the increasing usefulness of activities. E.g. suspending of unused sets of windows is something I’m really waiting for. For now I just ordered a memory upgrade for my laptop ;)

However I also see a problem with activities vs. VDs, especially for people not familiar with these concepts. With two switchers for VDs and activities. IMHO a more integrated concept is needed here. I don’t question that both have a right to exist. E.g. a real world example: cooking. The activity is “cooking”. But when you cook something more sophisticated you need several tables i.e. “virtual desktops” to prepare all the stuff. Just the integration between the two needs to improve.

Something I remember from enlightenment is that it also had two different kinds of VDs. With different ways and animations for switching. On the one hand there were desktop layers stacked on top of each other. And on the other hand each of these stacked desktops had several screens horizontally.
In our case the stacked desktops could be activities, while the horizontals would be the available VDs for each activity. Something like this might give a more coherent concept.

Some technical questions:

– What about applications with several mainwindows like Konqueror or Konsole? How does suspending and restoring of individual mainwindows work? I’m mean, it seems you use normal session save and restore, which usually applies to entire applications.

– Currently restoring of activities with several apps/windows can take quite some time, I guess. Especially on low end hardware. What would be extra cool: really suspend windows/apps by writing some image to disk. This probably needs support from the OS. But it would make the whole process more seamless and faster.

Jeffery MacEachern says:

“What would be extra cool: really suspend windows/apps by writing some image to disk. This probably needs support from the OS. But it would make the whole process more seamless and faster.”
I have heard that such a thing exists on Linux, at least (i.e. “single-app-hibernate”, if you will), but I know nothing about it, and I think when I brought it up with someone, the response was that lack of a (sane) cross-platform method of doing it would likely be a deal-breaker. It might be worth asking someone who actually knows about that side of things, though, as it’s an appealing concept.

Chani says:

ooh, I like the cooking analigy. :)

on applications with multiple windows: konqueror has a setting for whether to have several processes or one; I don’t know about konsole. Anyways, windows that belong to the same process aren’t shut down until that process isn’t on *any* open activities; so yes, you might have windows running but inaccessible (until you reopen the activity). I’m hoping to reduce this issue in 4.7, but it requires changes in both ksmserver and applications, and probably won’t be trivial.

I am a late commenter, but I just wanted to add my $0.02.

I haven’t really found activities super useful so far, but the fault doesn’t really lie with KDE (one of my main activities involves not using KDE as KDE is too bulky and I have limited memory). However from the screencast I’m really excited in trying it out again.

One time where I really found activities extremely useful was when I had to share my computer with another person. Now _that_ is something a virtual desktop really can’t compare!

arwa says:

I do not understand. How were activities useful in this situation? On my PC different users use different accounts and so can configure there own KDE (or whatever) the way they want. The can use multiple or only one VDs and activities, etc.

Chani says:

yes, for regular users accounts are best – but guests usually aren’t worth that much effort when they just want to borrow a web browser for a minute. :)

I hadn’t thought of using an activity for that; it’s a good idea. :)

arwa says:

I still do not see the point. Why not offer the guest just a blank desktop? Where is the advantage to offer him a blank desktop with another activity?

Comments are closed.

et cetera
%d bloggers like this: