Uselessd - stripped version of systemd
 

What do you think about this new project?
A valid alternative for slackware?

Interesting. I trust Patrick to make decisions about Slackware.

Having read the web page, I fail to see the aim of this new project. Beyond being fun for its developer, which is a perfectly legitimate goal.

An alternative to what? I don't see the need for replacing Slackware's current init system, that works.

On the other hand, if in the future Slackware has to ship something looking like systemd (most probably then because upstream developers of desktops would make that a requirement for future versions), we would need other components than that just the init sub system to be replaced, and all the parts will have to fit together - which is at the core of our BDFL's effort, I guess ;)

In any case this project will need a few years to mature and the situation will have changed then.

Conclusion: wait and see.

Needs more frop
 

In unreceptive viewers there is a danger of synaptic backfire in the brain pan
caused by overheating of the attention gland. This dense packing of brain cells
without proper organic coolants, such as frop [...]

>uselessd (http://uselessd.darknedgy.net)
>darknedgy

Yeah, no.

Already an existing topic:

http://www.linuxquestions.org/questi...rg-4175519475/

I have mixed feelings on it as my other reply states. Yes, journald's exoticism was pivotal, but they removed logind which is being used by the systemd-shim projects in both GNU/Linux and the BSDs. So not the best maneuver, plus the website fringes on blatant insult humor rather than being one of seriousness. It's a serious project granted, but it's masked in such as abstractness, you'd think it was just a joke.

I'm actually seconding Didier's comment: "Wait and see."

But in my opinion, more seriousness could have been a better approach. However, I do admire his stance about Distribution maintainers getting a bad case of laziness.

First of all, kudos for funny :) The systemd "discussion" often lacks humor.

Presently, it has 2 or 3 deal-breakers for me on alt distros I have that use systemd. BUT, it is perhaps an interesting "proof of concept".

It concept is sound. Strip it down to the bare essentials and go back to the bazaar methodology.

The author has many sound concepts as well once you get past that ghastly opening statement. It really rips at a lot people for various reasons, but once you read on, you find that nothing listed is based in any bias.

It would seem to be of no benefit to Slackware, since it only provides init and we already have a nicely working init.

What Slackware is lacking is reliable and well maintained alternatives to the non-init parts of systemd (logind, udev, etc.). Many components already depend on them (and likely more in the future). This is where the problem lies, so from a Slackware perspective uselessd is exactly that, useless.

Well, I'll say one thing about uselessd: reading over its intro page has been far more informative and succinct about what systemd contains than any hot-headed blog post I have seen anywhere else on the 'net.

I am still unclear about the real advantages of systemd. I get that it is super advanced and nifty, and I do use it on some spare boxen, but honestly as a user I have not ever looked at Slack's init system and felt that it was insufficient for anything I ever do. And I don't think I'm a light user of my computers; I use them, I use 'em hard. And Slack's init system works. So, no rush for me to get to systemd, nor to a replacement for systemd.

Even so, the existance of uselessd is nice to know about. Open source. Kinda neat that way.

Are there any alternatives to this components: logind, udev, etc.
Other important point, most framework depend or will depend on systemd (including KDE).
Knowing this makes me wonder if a bloated free version of systemd does not have its merits.

There is systembsd (git repository) by Ian Kremlin/Sutton, which might be able to provide some of systemd's functions.

edit: I've just read the comments: it seems systembsd is quite BSD-specific, so it needs to be ported to Linux. There was also some talk of porting it to a minimal busybox-like system, stripping out glib.

For udev, you have the likes of eudev and mdev. systembsd attempts to replace many of the other parts (like logind). The problem with eudev and systembsd is that they are relatively new projects and have not (yet) established themselves. mdev is part of busybox and has not really been used for the same types of systems as udev (i.e. desktop and servers), so it is not yet clear that it is up to the task.

In summary we don't yet know if we can rely on the various alternatives to provide all that is needed and in addition, that the various projects will continue to be well maintained. More work is required followed by a leap of faith that these are better than just taking on systemd, warts and all.

I'm not certain of what you mean by "framework" in this instance but if you are talking about desktop environments depending (or potentially depending) on systemd in the future it is most likely because of things like logind. The init component (which is all uselessd provides) does not matter as far as I can see.

So it has merit only if you want a new init for its own sake and as far as I can tell, the answer is that Pat doesn't want a new init. He has almost said this when ReaperX7 once suggested considering OpenRC instead of systemd.

Or to put it another way, we will only get a new init when the current one cannot be used.

This is most likely to happen because of the non init parts of systemd. Leaving two possible scenarios in my eyes. We use replacements for the other systemd components in conjunction with the current init, or we switch to systemd.

Alternative inits such as OpenRC, Runit, uselessd, etc. are pure fantasy and are not looking likely any time soon.

P.S. All these posts and no ranting and raving thus far. Best systemd thread we have ever had on the Slackware forums. Or am I pushing my luck saying this out loud? ;)

For udev, there is eudev.

And for me the slackware init, which as I understand it is actually a bit of a sysv/bsd hybrid, is still doing the job just fine, and I am not a fan of fixing what is not broken.

That said, if there is a need for a change going forward, I rather like the init posted here (near the bottom in a blockquote.)