It would be odd if those on this forum and KDE developers were not aware of the extreme anxiety among end users about the security of using KDE and all its applications in the foreseeable future.

This anxiety discourages the adoption of Free Software because people fear useful applications will cease to be available or will become unmaintained very soon because of incompatible dependencies.

One of the great Free Software principles was that applications should be portable, so anyone could use them in a platform-agnostic way, thus maximising utility. However, there is now great anxiety among many users that they will be unable to use the best applications because they will be made dependent on platforms that are not friendly to maintain.

GNOME is already dependent on systemd and therefore unavailable to all other users, and it seems KDE now plans to go the same way. I know nothing of the technicalities, but like many others, am alarmed that I will soon lose all the KDE apps on which my computing experience depends.

Must I return to Windows or Mac? Is Free Software a failed experiment? How can I have confidence my data will remain accessible and my applications will continue to be available on my chosen distro? The only information I've seen suggests distros have 6 months to adopt systemd or lose KDE. Many of us fear both choices and can see no way forward.

What can KDE do to reassure us it's worthwhile carrying on?

1) 6 months as a drop dead point - this assumes:
- all distros make Plasma 5 mandatory
- all distros stop supporting existing versions with KDE 4 immediately
- no one forks KDE 4
- no one forks Plasma 5 to not use systemd
Obviously 6 months is not a point where you are at risk or have any reason to worry you just won't have the latest versions

2) There are still a multitude of alternative wm/de's that are shipping, why would you not be able to use the majority of KDE apps under them? This is a question - I don't the answer but guessing that for most apps you would

3) If the vast majority of distros migrate to systemd why do you believe the few holdouts are correct and that Fedora, Debian, the *buntu's, openSUSE, Arch, etc are wrong. Maybe the discussion should take place on PCLinuxOS's site as to why they aren't planning on moving to systemd.

4) If RedHat and SUSE use systemd and they are the 2 larger commercial distributions and if there was the risk you believe exists wouldn't their commercial customers be leaving/screaming/switching

ps - I have no dog in this hunt, personally I don't care my system has systemd and it just works

Plasma does not depend on systemd and there are currently no plans to introduce a hard dependency. We use useful features provided by systemd and logind and you will miss those features if you don't have systemd or logind. That's the same as with all other optional feature. In addition almost all developers are using systemd and/or logind, which turns the alternative ways basically unmaintained.

The only way to ensure that Plasma keeps working without logind/systemd is to start hacking and maintain those code paths. I for example asked for help in the logind-kwin integration, but didn't get any help so far. If at all agressive comments from the "systemd haters".

I think you are targeting different possible topic:

• you are not forced to stop using your applications. Programs like Digikam, Okular, Amarok, Calligra Suite, and so on, can be used on every desktop environment (just like you can use GIMP, Inkscape, Firefox, Evince, etc etc). No risk of losing your favourite applications. Many of them are also available on different platforms, so I don't see which risks are we talking about here (apart from lack of developers, but that's a different issues which touches FLOSS and non-FLOSS applications).
• systemd (which is still free software) dependencies could be needed (but *I'm not part of Plasma team*, so please do not quote me on this) by some components in *Plasma* only. Which is just one of the many Free Software and content coming from the KDE community. Now, the reason for requesting some external, system level components to do some of the job (especially in session management) may be legitimate. The point here is what component should do this job. Luckily there are projects to provide alternative implementation for those components (for example, alternative logind).

Anyway, the world is not exploding and Free Software is not a failure.

Disclaimer: I am not a developer

I have not a clue what insecurity you are talking about, really. Why on earth would a system using systemd be less accessible? Could you please give some well funded and reasonable sources on where you read such a statement?

Since you admit yourself not knowing the technicalities behind systemd or even the basic system, what do you worry about? All these discussions about systemd or not systemd are purely technical, the very emotional parts come with extremely little technically funded reasoning AFAICT.

Seriously, either give some well funded reasons why a system would risk to make data inaccessible due to systemd or just use your computer and trust the developers! I am pretty sure that Linux has no risk whatsoever to fall into pieces

Thanks for all the responses. However, I don't quite understand all the comments. Surely, if distros want to stay with Plasma 4 they will be staying with unmaintained code and vulnerable to security issues, or will there continue to be security patches for Plasma 4?

I didn't suggest systemd is a threat to data. I meant if applications stop working because systemd is not available on the platform concerned, data might become inaccessible.

Although applications like Okula are usable on other desktops, I assumed that was because the KDE libraries on which they depend can still be present irrespective of the DE in use. Are we wrong, then to think those libraries will require systemd? If so, communication about how systemd affects software has either been lacking or not understood..

I chose the title of this thread carefully. Users are worried by a change they don't fully understand but which seems to have threatening implications. It matters little whether the threat is real. What threatens the adoption of Free Software is the perception things might become difficult to live with, and the lack of reassurance about alternative ways forward. "Systemd haters" is probably a bit strong, but a change of this scale is likely to meet resistance, at least until it has proved itself, from people who are naturally cautious, and those cautious people will perceive a threat to the whole infrastructure if an article (https://soylentnews.org/article.pl?sid=15/02/20/101235) is published suggesting systemd will become an essential dependency within a shorter time than they (or the platform they use) can come to trust it.

From what mgraesslin wrote above, it sounds as if the dependency on systemd has been exaggerated in that article, and the resulting anxiety may well be a false alarm. If that is the case, perhaps a clarification is urgently needed.

kjpetrie wrote:Although applications like Okula are usable on other desktops, I assumed that was because the KDE libraries on which they depend can still be present irrespective of the DE in use. Are we wrong, then to think those libraries will require systemd? If so, communication about how systemd affects software has either been lacking or not understood.

Our libraries also support Windows and OSX, given that I consider it as highly unlikely that any has a mandatory dependency on systemd. Also why should an application like Okular depend on systemd?

The mentioned article is exaggerated as even if we drop the non-logind code path in the Plasma workspace (not Plasma, not applications) it's not a hard dependency on systemd. It's a dependency on a D-Bus interface which any application can provide. I encurage any concerned user/developer to write a small wrapper which maps the dbus call to whatever is used on the non-systemd environment. This makes our maintenance much easier. If you read David's blog post you would understand why we as developers have an interest in using the D-Bus interface instead of having a multitude of many different ways which nobody tests.