How many times does the unix way have to be defined and for how many more decades? RTFM.



You mean you have not heard or seen the answers you want to or like. The argument against systemD is what is called Emperical Evidence. There is more facts, witnesses, against (even in this thread alone but staggering scattered around the net) systemD then for it. Just because people refuse to consider the emperical evidence, do not understand the term I just used, or just do not care, does not change one iota what is stacked against systemD.

2 members found this post helpful.

Exactly. Slackware has unique init that no one else has. The differences may be small compared to similar inits (and talking in the whole) but it actually makes big differences in the real world. systemV always sucked but it was what people had. If you dig you will find all kinds of opinions going back to the early '80s attesting my assertion. It is just not my opinion.


Instead of adopting Slackware' s init, OpenRC (I quite like it), runit (my fav), a BSD (again they differ between them), and getting rid of systemV the elitists in the Linux world went nutty and pushed systemD upon us.



Oh well. It will die.

2 members found this post helpful.

Hmmm, no rude behaviour? Perhaps you have a different meaning of "rude behaviour". Seems Richard Cranium called it right. And that's "systemd", not "systemD" BTW, perhaps you need to RTFM.

Seems that way, I'm out now too...

Hmm. I write it systemD when I feel like it because of my total lack of respect for it and the developers complete lack of respect for me and others.

Post one example of rude behaviour. Not from me, because many people interpret my directness as rude and if they do too bad. As my comment was not directed at myself and in general please provide one example.

Sorry, but the number of people who are against systemd is not empirical evidence of anything but the state of public opinion. There are some real arguments against systemd that have been aired in this thread, but how many people vote on either side is totally irrelevant. After all, there was a time when most people thought the earth was flat (though educated people like Dante always knew it was round). Now practically everyone agrees that it is round. Does that mean that the actual shape of the earth has changed over the centuries? Of course not!

Let's stick to arguments about how systemd works (or doesn't) and leave the numbers game out of it.

4 members found this post helpful.

With all due honesty how well it does or doesn't work is not the only issue, the development standards and developers behind the software are a large contributing factor to why many people are not interested in using it. You may not deal with developers, but unreasonable or hostile developers are a real concern for package maintainers and other people reporting or fixing issues.

It is the developers that lead to current situation where systemd has a large amount of technical debt, it has many different moving parts and a lot of lines of code with great potential for current and latent bugs. The increased complexity also greatly increases the burden and difficulty of fixing any bugs.

For example a while ago I was dealing with a niche udev issue until I discovered it was fixed in the latest udev releases in both systemd and eudev which Slackware uses. Since Slackware uses eudev it was easily fixed upon reporting the issue and solution where Pat updated one package in current. If Slackware was using systemd fixing this would of not been nearly so easy given the increased complexity and technical debt behind integrating an entire ecosystem into PID 1...

By all means systemd is not the only program that has faced issues like this, wise developers are unfortunately a rare thing...

5 members found this post helpful.

Yup.

Udev is necessary, and we had that long before systemd. We had to switch to the forked implementation because of the kernel hooks. I don't use consolekit, policykit or udisks2. I don't have to have those things on Slackware. Poettering stated a goal to make components of systemd inseparable so you couldn't do things like that, but thanks to Gentoo devs and eudev, we can still function.

4 members found this post helpful.

You are aware that your own story is a proof against your own words?

What happened, in fact?

1. You are dealing with a niche issue of UDEV
2. You are not alone on dealing with this niche issue on UDEV, and someone else reported this bug to the systemd developers
3. The systemd developers, over this "bloated" source code fixed that particular bug and they made a new release
4. The Gentoo guy grabbed the new systemd release and as usual with his machete extracted the UDEV, in a new proud eudev release
5. The Slackware updated the EUDEV to the new release, and that made you happy

I do not know others, but I do not believe that we should be proud, waiting like some puppies for someone else to give us a fresh hulk of systemd.

This is the real issue.

2 members found this post helpful.

This UDEV "fork" is today just code chopped from systemd by a Gentoo guy using a machete. I will not be sooo proud of it.

Me neither! Though I can see some advantages in principle from policykit, compared with sudo. What consolekit is for, I never worked out.

Someone stating a real world event on how systemD broke or mangled their system, is a piece of empirical evidence. Yes, the number game comes into play at some point if it is an 'intelligent assesment'. At some point one has to count the pros and cons of an argument and the amount of instances of arguments that have been observed. Sooner or later it always involves numbers in some way or another.



You use the the example of flat earth and have no proof of your claim. That is not empirical. For it to be empirical there has to be textual witnesses (any living witnesses to your claim are all long dead). Well, there is nobody in texts that claim the earth was flat as your FUD example you learnt from someone else and never checked yourself claim. BTW, flat earthers do not claim a flat earth but Flat Plane. You just provided an example of an opinion vs empirical evidence. You cannot provide me one quote from an ancient or medieval text that such a claim was made and claim everybody believed it, which is hogwash. There was many thoughts and ideas floating around for thousands of years, and the predominant one was a Flat Planar. Also, it was Copernicus that brang the Heliocentric model in and became standard and Dante had not one little thing to do with it. Show me in Dante' s work he thought it was round? I have read his poems. The globe earth is from the Heliocentric model, and has stuck ever since that became a standard model.

In short, yes, the amount of people against systemD because of experience with it IS Empirical Evidence and the numbers do count. You do not build a bridge by discounting the numbers of factors that cause failures, their chance of failures, etc. You count them into the equation because otherwise it is not engineering, it is not technical, it is just plain ignorance.

2 members found this post helpful.

Its alarming that some devs could just decide to replace all those things in one monolithic package; with the stated purpose of merging components you plan to replace into one inseparable package; and then some how getting the kernel devs to change what got changed (in this case udev) to effectively (if not intentionally) force adoption along. If nothing else I have to say thank you to Poettering for the wake up call.

7 members found this post helpful.

Well, this is rather off the main point of the thread, but if you read the Inferno, you will see that hell is presented as a funnel-shaped volume under the surface of the earth with its point at the earth's centre. Each circle is smaller and more crowded than the one above. At the bottom of the funnel, Lucifer lies entombed in a block of ice. Dante and Virgil climb down to his waist and then up his legs towards Mount Purgatory on the other side of the earth.

Dante asks why they are now going upwards when they haven't turned around, and Virgil explains that Lucifer's waist corresponds to the centre of the earth and that any direction away from this will be "up".

Of course Dante and his contemporaries believed that the spherical earth was motionless at the centre of the universe and Copernicus believed it went around the sun, but that is a different issue altogether

2 members found this post helpful.

You're the one posturing, parroting and making semantic arguments. Why would I be "proud" of it? I didn't write it, but I do need udev. Spare me the pedantry... I know it's systemd's udev implementation.