Interesting thread over at dyne.org

Originally Posted by hazel Why is it that any thread which has to do directly or indirectly with systemd always ends in acrimony?

It's an argument thread. It may be better on Debian, or general, where most argument threads are found. The USA seems very bipolar since Trump got into politics.

But of course, our BDFL said no to Debian, which might be why it's here. I think LP moved to M$ at exactly the right time, when systemd was at it's peak. It can only go downhill from here. Remember Selinux?

Originally Posted by hazel Why is it that any thread which has to do directly or indirectly with systemd always ends in acrimony?

Perhaps because the introduction and adoption of systemd strikes to the heart of the philosophical question of the means justifying the ends.
For many, the adaption of the startup initially introduced by Apple OS to produce fast boot times, was seen as an inevitable evolution. The advantages in provisioning virtual machines in the cloud were irresistible.
This came at the cost of the violation of the UNIX principle of ‘a program should do one job and do it well’.

Honestly the specifics of systemd and UNIX principles are I think a distraction. There are use cases like cloud provisioning for which systemd or something like it certainly offer advantages over the things that came before. It is not the only way to go about getting faster boot times, on user facing machines.

I would argue if you are booting your PC very often you are doing it wrong anyway - why wouldn't you want to generally preserve state with suspend or suspend-to-disk?

It is far from the only way to deal with things like service recovery on servers; but it isn't a bad way. the need is obsoleted for large scale application by containerization on cluster control like k8s anyway and recovering at the container level anyway.

What everyone has forgotten about the UNIX world and this even true of the decision to use UNIX itself is that everyone has various problems. Some problems looked enough like the problems other people had had before that they decided they could share stuff. It was a highly diverse very DIY environment where everyone rolled-their-own and if it was good other picked it up. What makes people mad about systemd, is the fact it did not gain ascendance by a bunch of distro maintainers independently going "wow that looks way better than my copy of init and the shell scripts I have now, I am switching!" It was and has been a corporate decision to pressure everyone else to standardize on it, by exerting influence on other FOSS projects to tightly integrate it.

The strong feelings are not about the tech, its about the arm twisting and revelation that a lot of really key in things in the FOSS "community" are really very much yolked to a few corporate masters.

Originally Posted by _blackhole_ Hello there! I am not the thread starter. The thread probably needs to be moved to general? (As a "FreeBSD user", I am somehow not entitled to post here [in the Slackware sub forum]!?)

You're always welcome to post here.

One of the things I like about Slackware is the pragmatism displayed in its development. Another thing I like is the (apparent) adherence to the 'principle of least surprise' as development moves forward.

I'm confident that PV will find a route through the /usr merge and will navigate a way forward that maximises our freedom to 'mix and match' as Hazel put it up the thread.

Best wishes to all.

Originally Posted by _blackhole_ Hello there! I am not the thread starter. The thread probably needs to be moved to general? (As a "FreeBSD user", I am somehow not entitled to post here [in the Slackware sub forum]!?)

You are always welcome everywhere

Everyone,
Not sure how the threads with systemd always devolves but one thing is for sure... It works for M$ in the enterprise market.
Couple of weeks ago at work I updated the "users" file for radius auth in a RH server and had to use the systemd commands to restart the service. Next thing you know, the Security teams rocks up and double checks what I was doing on the radius server.
They only use winblows and some M$ Server geared for security flagged what I did. I'll ask what the M$ server is runing next time I'm with that team but don't hold your breath

The ad hominem attacks against people who take issue with systemd for technical reasons or are opposed to anti-competitive conflict of interest really are the primary issue in these threads. People like @_blackhole_ are literally the obstacle to systemd adoption. Sure he is welcome to post here but that doesn't change the fact that he's a bit of a dumbass when it comes to the reasons behind all the complaints towards systemd. They see negativity and immediately feel an overpowering urge to attack it without taking the time to understand why people got so angry in the first place.

Artix provides a much better explanation for why they forked. Like Devuan it's primarily due to systemd, but they put more effort into collecting all the various examples into one big FAQ. They even rely on First Person references.
https://artixlinux.org/faq.php


Lennart is responsible for what Lennart says, and what Lennart does. And if Lennart is going to be a whiney ass little dipshit he will be treated accordingly. If @_blackhole_ can't wrap his head around that concept then IMHO he is not welcome. Lucky for him I carry no authority on this forum so he can safely disregard my opinion.



^ and THAT was the point of my epic one liner about freebsd/devuan/slackware/windows users... The chaotic stupidity of random semi-anonymous users who opine on a subject based on raw emotion rather than emperical data.

Originally Posted by Pithium The ad hominem attacks against people who take issue with systemd for technical reasons or are opposed to anti-competitive conflict of interest really are the primary issue in these threads.

Exactly. There is no discussion allowed. This issue is so divisive... almost like it was engineered to be that way.

           )
       '
   (    ',   )
    '  )  ( .
 ).   (  ,   )
(_,).. )  . (' )
^^^^^^^^^^^^^^^^
|              |
|              |
<==============>
|              |
|            ()|
|              |
<==============>
|              |
|              |
^^^^^^^^^^^^^^^^

Yet another dumpster fire about <"solution"> to a problem no one really had.

In the beginning the struggle was about to prove Linux/GNU was viable for profit.
Then came the time it was tested for just that - and RH lead the way - more or less successfully and exemplary...
Then, once proved viable, there came people to try free-ride or profit off of it (or both).
Then came people who were crying out foul for just that...

Tl;DR:
I guess

"You can only fool some people all of the time, but it is just enough for a decent living"

does apply here?

FWIW every time i come across events like that (frankly a phenomena as old as man kind) i come to appreciate Slackware a tad more.

@PV, sire, my hat is off to You and all the passengers in this little Boat of sanity we call Slackware

As withe several other posters in this thread I too have come to SlackWare after twenty-three + some odd years of using pay to play Linux like Red-had,Zorin Ubuntu and others. I really hate to admit that I too have come from closed source operating systems/platforms (i.e.Windows).

My introduction to GNU/Linux came in the form of a downloaded copy of Debian this of course was after I obtained a boot about RedHat Linux, I can still vividly remember having to step through several hours of debugging failed build of the proprietary Windows drivers for BroadCom WiFi adapters via the use of the B43-cutter package failing to properly extract the drivers for aforementioned Windows and rebuild it to play nicely with the Linux distribution that i was using at that time where based on Debian Ubuntu for example.

I've grown sick and tired of the literal erosion of the non-systemd Linux distributions, over the last few years I've bounced around between both systemd and nonsystemd based distributions I was ready to swap from Debian to Antix or was it Artix I did a quick reconsider my thoughts and come back to SlackWare after having given up on it on several earlier occasions.

Originally Posted by niceflipper8827 My introduction to GNU/Linux came in the form of a downloaded copy of Debian this of course was after I obtained a boot about RedHat Linux, I can still vividly remember having to step through several hours of debugging failed build of the proprietary Windows drivers for BroadCom WiFi adapters via the use of the B43-cutter package failing to properly extract the drivers for aforementioned Windows and rebuild it to play nicely with the Linux distribution that i was using at that time where based on Debian Ubuntu for example.

You had bad luck. My previous laptop had a Broadcom card. I initially installed Mint on it because I knew that Mint has a reputation for being friendly to exotic hardware. It has a package called fw-cutter which can cut out wifi firmware from the driver packages of several chips including Broadcom's B43. It worked out of the box and I was able to use the resultant firmware files in any other distro I wanted.

Originally Posted by _blackhole_ This is true, but the "support" is from Debian - i.e. the project providing the security patching upstream: https://www.debian.org/security/ The derivatives, such as Devuan and the other smaller Debian based projects, have none of this infrastructure. As with any other derivative, Devuan simply has a "fan base", who are mostly ideologically driven, who post on a forum or community and those maintaining the distribution are usually just "packagers". That's probably somewhat of a generalisation, but in a nutshell it sounds about right. Looking at the linked dyne.org thread in the OP there is the usual hysteria, misinformation and hatred of Poettering and about systemd supposedly being a Microsoft backed project all along. Someone posted a conspiracy there about MS, IBM and Red Hat attacking and destroying the Debian derivatives through Debian itself... This person clearly is misinformed as he equates systemd to the Windows registry - this is a a myth I have seen before, spread unchallenged in certain circles - usually by users of Devuan. There is another equally flawed comparison to Windows svchost, it goes on and on - in essence criticism of something they fear, with no technical data to back that up. Their forums are littered with similar, usually worse, crap and it's precisely why the project isn't take seriously. Since it was founded, they've done a great job of driving people away - never anywhere near a serious threat to Debian (or systemd). More likely the truth is, that none of the aforementioned fortune 500 companies actually care about a Linux distribution used by < 0.01% of Linux users. These hysterical misinformed types have only created this "assault" and conspiracy in their own minds. Debian went down the systemd root, because that's what enterprise server market wanted, and those on the corporate payroll have been steering things for well over a decade now. The enterprise market is the only Linux distro usage that actually registers on the scale. The rest of Linux is Android and embedded use. Desktop hobbyists are tiny. Linux got too big, too popular and has evolved accordingly. Some of that evolution is good, some is bad. The problem with corporate backed software, is that it's developed from a business perspective - ultimately to make money. For Red Hat, a system which is much more black box and needs accreditation to administer by their "certified professionals" and is dependent on their paid support contracts is simply good business.

All I've seen in your posts is pure speculation and guesses pulled from the nether world with nothing to back them up. "Devuan simply has a "fan base", who are mostly ideologically driven, who post on a forum or community and those maintaining the distribution are usually just "packagers"...you say it's a generalization yet still stick by it as what should be believed. Then again, I could pick out this kind of "generalization" that you paste on 'Devuan' users and 'Devuan' dev's and the OS itself, and again though it's all speculation and just plain "I think this is true and so should everyone else" type of bourgeoisie claptrap from someone who thinks they know better than anyone else. It seems to me, that those in the know, know far better than you what will be best for systemd-less OS's. If it was so good, systemd, why hasn't Linus Torvalds embraced it? Why do so many others with far more knowledge of what it is and what it does and what it *will* do to Linux, than you have about it (I'm presuming), embraced it also? You know, the ones that actually count and who *are* devs for systems and programs? I also believe you are very wrong about the folks who left Debian to start Devuan...they are also the 'devs' who work on the OS and worked on Debian. They knew something was wrong with it and weren't just something like testers for Debian, they know what Debian is and worked on it and chose to step away from the oncoming screwup that systemd was going to be and has become.

If you want more people to read your posts and not be put off by them, how about giving what you put in them a little actual *fact(s)* over all the speculation and guessing. You know, some cites and actual percentages with proof of them for the generalizations you spread so incessantly.

Originally Posted by FTIO If it was so good, systemd, why hasn't Linus Torvalds embraced it?

I suspect it's partly because he once had a bitter quarrel with Kay Sievers that ran to a great deal of bad language. Afair he had two particular grievances:

1) Systemd crashes if your kernel doesn't use cgroups because some pointer is then given a null value. Linus thinks that no pointer should ever have a null value because the program should always test for that and exit gracefully with an explanatory warning. Especially if it's the init program! Sievers thought that it was entirely the user's fault if systemd crashed because "everyone knows that systemd needs cgroups".
2) If a developer runs the kernel in debug mode to test new code, systemd automatically runs verbosely too. That means that kernel boot messages can get drowned out.

All this ended with Sievers getting booted off the kernel team.

https://www.theregister.com/2014/04/...evers_dust_up/
https://www.reddit.com/r/linux/comme...rs_from_linux/

Duplicate.

Originally Posted by hazel I suspect it's partly because he once had a bitter quarrel with Kay Sievers that ran to a great deal of bad language. Afair he had two particular grievances:

Is there still a language issue on the LKML or has that been toned down? Incidentally, I'm leaning towards Linus on this one. His objections are usually valid enough, from the reports I've come across.