It is possible thanks to patches, but more and more packages requiers to have udev built with "systemd" USE flag.
It's much more easy to switch to eudev than write more and more patches to prevent systemd dependency.

1 members found this post helpful.

But not, for example, openwrt.
https://bugs.freedesktop.org/show_bug.cgi?id=73729

Finding ways to get device driver maintainers to just unilaterally switch coding from netlink to kdbus is going to be like trying to herd mice. Even as of 3.18rc4 it's still not included. The patch was submitted but still no word yet on if it will be used, and even then, if no device drivers are redrafted around kdbus, how useful will it even be if udev scraps netlink and ironically breaks systemd?

It's ridiculous having to visit a site called "freedesktop.org", if upstream broke something in your embedded distribution.

Embedded developers enjoyed the ride, now it's time to fork.

Jesus Christ, Lennart answer made me angry.

After reading all those replies, it only seem that Kay and Lennart both don't care about Linux itself if you read it right. All they seem to care about is systemd, and the rest be damned. Plus a lot of Lennart's replies are more trollish than actual technical talk. Twisting an argument to say he's racist, asking someone about their puberty, etc. Talk about cheap and low brow. If you said anything like that on a public forum, you'd be banned without hesitation.

It would appear that they're trying to lock out other libcs even from udev now, and have done so. Glibc isn't the only libc that has ever been used for Linux systems. Many of these other libc are smaller sized variants that cater to embedded projects, such as OpenWRT where ROM size is limited and important. You can't always have a full sized package on a ROM, and need a smaller or micro implementation. If you have a router manufacturer that use OpenWRT, or similar, ROM chips costs manufacturers money, and the more space is required on the chip... and the larger that chip, the larger the costs both to manufacturer and consumer alike.

Hmmm... I honestly wonder now how much time is it going to take before all non-systemd systems are forced out of udev to alternatives like BusyBox's mdev and such? From what I've read into mdev, it requires a bit more work to get running and a lot of extra configuration, but in the end, could it be worth it?

Doesn't OpenWRT have hotplug2 still in development or was that project dropped?

But not from a Red Hat site (what freedesktop.org is). There only criticizing their mighty employer gets you banned. Totally unprofessional. Still wondering why the "Free Desktop" was a total market failure?

To be fair, while Lennart and Kay were completely dismissive of someone outside their own designated use cases (as they always are), it was a third party (not the bug submitter nor the developers) who usurped the conversation and he obviously came in with a chip on his shoulder. If he had been less antagonistic, then perhaps a solution could have been reached, especially since the missing %m support from uclibc was a deliberate decision on his part to compile uclibc without it when it is supported. There may be a bug with --disable-localed (I'm not going to try to compile it to find out) that probably could have been fixed but it will likely be ignored now because of the nastiness of that bug report (unless another bug report with more details and an actual patch is opened without all the hostility). So if there is a bug, and that bug was fixed, it may compile against uclibc...we just don't know. But people like to jump to conclusions.

1 members found this post helpful.

Very unprofessional. And no, I know it's a total failure, many people have said GNU/Linux can't compete on the desktop, and now I see why. It's people like Lennart and crew who can't get their act together and fix bugs in software, but insist on pushing out feature after feature until the project implodes that they finally learn something.

This reminds me exactly of what happened to the emulator ZSNES. They kept pushing out more and more video filters trying to add everything under the sun to make things look nice, while bugs and long known workaround hacks, rather than trying to implement actual emulation of hardware, were slowly eating away at the project's ability to actually emulate an SNES successfully much less run games at all. In the end, ZSNES imploded after the developers got tired of people pestering them endlessly to fix bugs, yanked all new developments offline to a private subversion network, and relicensed the project under a proprietary license rather than the GPL license it was under for the longest time. Now ZSNES is history.

I'm wondering if, yet once again, time is flowing like a river and history is repeating itself, but this time with yet another project?

Can you please back that up with clear and unambiguous evidence?

Of course, let me quote Wikipedia for you:

I find it almost a Shakespearian comedy that freedesktop.org was founded towards a POSIX manifesto, when it's efforts towards POSIX promotion have been more than lackluster.

Imagine you are a small startup with an OS based on a hobbyist kernel facing powerful UNIX vendors and free BSDs. You would totally be into "open standards" and "interoperability"? Wouldn't you?

Meanwhile you gained control over a huge part of the ecosystem and most of your competitors are out of business. So why should you care about higher goals now?

History of Google is basically the same thing: "Open web" bla bla for a few years including supporting niche browsers like Mozilla and Opera and now distributing their own version of Internet Explorer 38.0 with the web being in exactly the same situation as 1998: "Best viewed with Google Chrome". No, worse: Now, thanks to Mozilla failing, we have DRM in HTML!

I haven't ch3cked for Gentoo, since I currently don't have access to my machines, but the only patch used by Mr. Workman is one to provide more cd aliases:http://www.slackware.com/~rworkman/s...-from-systemd/
So it is not quite clear to me why Gentoo would need patches.

After reading this thread, how can you not indulge in fantasies involving the systemd developers and a very large bus?