You might have noticed that my yesterday’s blog post was an April Fool. With this post I want to explain a little bit about the idea of the article.
First of all I want to thank the VDG, especially Jens and Ken, for their support by providing the great mockups. That was kind of the first element to notice that it’s an April Fool: normally I do only integrate screenshots by myself and I would not have been able to show a drawing in Krita.
The idea for this April Fool is based on the reports about the inclusion of an adware tool called superfish on Lenovo notebooks. Especially the initial response by Lenovo claiming that this malicious software got installed “to improve the shopping experience”.
Nevertheless the article contains truth, half truth and thing we would never do. So let’s look at some of the ideas presented and evaluate them. The first idea presented was share buttons integrated in the browser’s window decoration. This is an idea which already got introduced similarily through Share-Like-Connect. Personally I think it’s a good idea if browsers would provide such global buttons and in return filter out all the user tracking share buttons on the web sites. A global area for share buttons directly integrated in the browser could increase the security of users a lot.
The next idea is about monitioring the usage behavior of the user. For this I must say that in a Wayland world KWin would have enough information to do that, but we don’t look at the information at all. We do not care what an application is: we don’t know that a given window is “Krita”, we don’t know that it should be used with a drawing tablet, we don’t analyze any input and just pass it through to the application. KWin is completely ignorant on what the application does with input events passed to them and as ignorant to what a window content looks like. To us a window is just a texture we render to the screen. We don’t know that there is a “drawing area” or a “video” element, it’s just one texture. Similarly we do not know (and care) about input events. All we do is determine which window should get the event and pass it to it. We are quite aware that by passing all input events through KWin we could do evil things. The good thing is that our software is open source and you can see what we do (relevant code file is called input.cpp, events come in from the files under the libinput/ directory).
Last but not least I want to point out that we as a window manager will not integrate advertisement into the window management. We do not care whether you watch a video, we might not even know it, because as explained: we are ignorant about what you do. I consider applications which integrate advertisement without the user’s explicit acknowledgement as malware.
The bad part of my blog post is that I expect these things to happen. And if we look around us we can see that it already happened: adware is unfortunately also available in FLOSS. And in the walled garden’s of the proprietary world we will see more attempts to gain more information about the user’s to provide even better matching advertisement. It will blur more and more.
Thank you for sharing your April Fool and the follow-up post. As usually, it was a fun read. Is there a plan for implementing DWD at some point? Given what I know, I think it is an attractive idea!
It needs people to work on it. I unfortunately do not have time to work on DWD.
Thanks for your articles 🙂
I (unfortunately) also agree with you about the “bad part” of it. As you mentioned, adware is growing everywhere and it is good that some projects like KDE (but also companies like Jolla) strongly commit against adware. Such a commitment is in my opinion even more important than open-source itself.
We need values and “fair” softwares and I am convinced that users will slowly understand that it is not a fatality and hopefully move to the right projects!
I thank you for this insight. I had grown accustomed to seeing the web like this outside my home network (where I block ads on the DNS level), seeing this in normal applications was quite a shock. I think the problem lies elsewhere though : Dancing pigs applied to “get something for free”. BTW where in FLOSS is adware?
Pigs will hopefully prove to have some kind of awareness 🙂
I guess he was talking about Ubuntu and its wonderful Amazon integration which, “of course”, is not an ad (ah ah). And/or about the amarok store plugin.
>We do not care whether you watch a video, we might not even know it, because as explained: we are ignorant about what you do.
Maybe you should though; it would be nice to be able to disable screensaver when a “big” application is doing alot of damage events or somesuch. It would be even nicer to decouple locking the session from blanking the screen, so that you could leave a video playing without giving full access to your computer.
And not only showing a password box when the session is locked and some input jiggled, maybe showing mpris controls too would be cool.
It’d be cool if pulseaudio also could block things from say putting the computer to sleep, say if an application is currently playing music. And if the screen blanker or session locker could send mpris2 pause calls and mute everything that didnt dissapear. And maybe choose policy per app. Stuff like that. That’d be sweet.
There is a dbus interface for applications to inhibit the screen saver/locker. It’s the task of the application to call it.
Inspired by your posts, I would like to share my thoughts.
I think that FOSS projects should more and more start to exploit the valuable information they can extract from their software to enhance it. They often lack manpower and have unclear directions. So why not to use information collected by the software itself to understand what is appreciated by the user and what is not? But often developers do not even know the extent of their user base. Isn’t it terrible?
I would create a free and public database collecting information like how many users OpenSuse has? How many Ubuntu? How many are using KDE, how many Gnome? Is Clementine preferred over Amarok? Is that specific shorcut used more than menu navigation or icon click? Those who installed Pidgin after how long uninstalled it? Did they try some other similar piece of software they weren’t satisfied with? Is that feature employed?
Information is the key of success. Google and others thought us that.
Don’t you agree?