LinuxQuestions.org
Visit Jeremy's Blog.
Go Back   LinuxQuestions.org > Forums > Linux Forums > Linux - Distributions > Slackware
User Name
Password
Slackware This Forum is for the discussion of Slackware Linux.

Notices


Closed Thread
  Search this Thread
Old 05-06-2013, 11:54 PM   #31
T3slider
Senior Member
 
Registered: Jul 2007
Distribution: Slackware64-14.1
Posts: 2,367

Rep: Reputation: 843Reputation: 843Reputation: 843Reputation: 843Reputation: 843Reputation: 843Reputation: 843

I guess I have an unfounded assumption that people should maintain a generally professional attitude here at LQ, and I am constantly surprised and disappointed to see people resort to cursing when they are otherwise unable to make a point....

Note: My commentary below may not suggest this, but I am not a fan of systemd. So keep that in mind.
Quote:
Originally Posted by ReaperX7 View Post
Systemd requires rebuilding new scripts to startup your systems completely from scratch, even though it says sysvinit scripts are compatible, but Lennart then retracts this in his manifesto and says they aren't.
The point is that all systems using systemd use the same init scripts (or blurbs or whatever they're called in systemd speak) so that one universal init script can be shipped in upstream sources for services. Thus, it does not require rebuilding new scripts to startup your systems completely from scratch at all.
Quote:
Originally Posted by ReaperX7 View Post
Removing the need of the shell? Linux and GNU is a shell based operating system, not a GUI driven point-click-and-go. What do think BASH, ZSH, etc. are for? To look pretty and take up space on the drive? Hardly. GNU/Linux is a command-line OS. It always has been.
No one said the shell would disappear and exaggerating a jump from point a to point b to make a point is disingenuous. The shell will be used as a userland utility, just as it is now, but will no longer be used in the init system. I am not a fan of this but to suggest that using systemd is an assault on bash is just ridiculous.
Quote:
Originally Posted by ReaperX7 View Post
And udev IS broken, every system maintainer, admin, and many users can attest to this, even those using systemd and even upstart. The rules it creates even on systemd using Linux distributions don't always work right either. This is the udev built into systemd and then stripped back out as it's own toolkit which is it, and it still has many issues that the developers haven't ironed out because they don't know how. Udev was a misguided concept that was a pipedream at best to replace DevFS which worked flawlessly using untested and broken code to create a semi-fluid system of daemon tools to manage the resources in real time, but it can't even write it's own rules correctly. How the hell does that work right if it can't work right at all?
It is nice to delude yourself into thinking that all was peachy before udev, but that doesn't really make it true. udev solved a legitimate problem. Whether it did it well, and more importantly, whether it continues to do it well, is another matter.
Quote:
Originally Posted by ReaperX7 View Post
If you have ever wondered why Slackware uses so much vanilla stuff, it's to AVOID endless patchworks and rhetorical bullshit the upstream creates as a backwash.
This sentence makes literally no sense whatsoever. Slackware uses vanilla upstream stuff to avoid crappy upstream stuff? Wha?
Quote:
Originally Posted by ReaperX7 View Post
Linux is a POSIX compliant UNIX-like kernel, and yes it's not UNIX, same goes for GNU, but it's foundations are in UNIX and BSD, and even HURD. Don't you dumb noobs ever learn history or do you brush that off like normal accepting what you want as history and tossing the rest in the trashcan?
Linux is not now nor has it ever been POSIX compliant. It is UNIX-like and that's it (though it is 'mostly' POSIX compliant). But I guess you never learned history or brushed it off like normal bla bla bla.
Quote:
Originally Posted by ReaperX7 View Post
Systemd spiltting across binaries means it's HIJACKING the system. When you have the shell, GUI, and APIs being controlled by a single entity, it means that if the core of the system fails, systemd in this case, the whole system crashes bringing down everything tied to it.
If init crashes it brings the system down. If a kernel module crashes it brings the system down (though obviously this is not tied to the init system). systemd is not one big binary; if part of systemd crashes, it will not bring down the whole system (unless that part is the core init binary [whatever it is called] or involves some other critical error...). Right now the entire system is controlled by init, so that fulfills your warped definition of everything being controlled by a single entity. I have no idea if you understand what hijacking means but while systemd may be hijacking the Linux community, it isn't hijacking any individual system.
Quote:
Originally Posted by ReaperX7 View Post
Doing one thing and doing it well, is a double-edged sword. Systemd does not do everything well. If it did, more developers and maintainers would sign onto it, but obviously, this isn't the case.
Given the number of distros that have moved to systemd, you appear to be making up statistics.

I hate systemd as much as the next guy, but spreading FUD and nonsense about it doesn't help anyone.
 
9 members found this post helpful.
Old 05-07-2013, 12:14 AM   #32
ReaperX7
LQ Guru
 
Registered: Jul 2011
Location: California
Distribution: Slackware64-15.0 Multilib
Posts: 6,554
Blog Entries: 15

Rep: Reputation: 2097Reputation: 2097Reputation: 2097Reputation: 2097Reputation: 2097Reputation: 2097Reputation: 2097Reputation: 2097Reputation: 2097Reputation: 2097Reputation: 2097
Yeah, but too much FUD as to it's usefulness is being spread around just as bad, if not worse. Again, I offer no apologies, but yes, udev did fix one problem, but it created a bigger mess that has never been cleaned up. All I offer is a wake up call to say, "Stop, Think, Backup, and Review before you jump off the diving board into a pool with no water."

Systemd might be revolutionary, but it's far from being a perfect replacement to the bsdinit and sysvinit startup script kits. If the only real fix it offers is a faster boot time, then realistically for a server environment, it's completely useless to a system administrator as scripts really only require a basic text editor like Nano or VIM to edit and boot times for servers are negligible. SysVInit and BSDInit can be scripted to work in parallel the same as systemd also.

It's basically reinventing the wheel and not making it any rounder.

Last edited by ReaperX7; 05-07-2013 at 12:23 AM.
 
Old 05-07-2013, 12:38 AM   #33
elvis4526
Member
 
Registered: Aug 2011
Posts: 114

Rep: Reputation: Disabled
Quote:
Originally Posted by ReaperX7 View Post
Read the articles posted at the front of the topic instead of shooting your mouth off like a jackass. You want to hammer at me like a troll, fine, I'll play your game. I've got nothing really to lose here.



Systemd requires rebuilding new scripts to startup your systems completely from scratch, even though it says sysvinit scripts are compatible, but Lennart then retracts this in his manifesto and says they aren't.

Removing the need of the shell? Linux and GNU is a shell based operating system, not a GUI driven point-click-and-go. What do think BASH, ZSH, etc. are for? To look pretty and take up space on the drive? Hardly. GNU/Linux is a command-line OS. It always has been.

And udev IS broken, every system maintainer, admin, and many users can attest to this, even those using systemd and even upstart. The rules it creates even on systemd using Linux distributions don't always work right either. This is the udev built into systemd and then stripped back out as it's own toolkit which is it, and it still has many issues that the developers haven't ironed out because they don't know how. Udev was a misguided concept that was a pipedream at best to replace DevFS which worked flawlessly using untested and broken code to create a semi-fluid system of daemon tools to manage the resources in real time, but it can't even write it's own rules correctly. How the hell does that work right if it can't work right at all?



You want to know how "fucked up" the "upstream" is? Go use ArchLinux and keep Pacman updated daily and see how long it takes before you have an unstable and unusable system. The upstream only cares about pushing out code flawed or not, and features that aren't tested. That why Patrick devotes so much time and effort into testing the shit out Slackware before upstream stuff comes along. If you have ever wondered why Slackware uses so much vanilla stuff, it's to AVOID endless patchworks and rhetorical bullshit the upstream creates as a backwash.

The udev in systemd has been extracted by various distributions because udev is required as part of the kernel resource management software itself, not the system daemons. Now that udev is merged into systemd, you have the kernel, systemd, and shells like Gnome so closely tied together, getting anything to work independently is a nightmare.



Linux is a POSIX compliant UNIX-like kernel, and yes it's not UNIX, same goes for GNU, but it's foundations are in UNIX and BSD, and even HURD. Don't you dumb noobs ever learn history or do you brush that off like normal accepting what you want as history and tossing the rest in the trashcan?

Systemd spiltting across binaries means it's HIJACKING the system. When you have the shell, GUI, and APIs being controlled by a single entity, it means that if the core of the system fails, systemd in this case, the whole system crashes bringing down everything tied to it.

Doing one thing and doing it well, is a double-edged sword. Systemd does not do everything well. If it did, more developers and maintainers would sign onto it, but obviously, this isn't the case.



Red Hat only proved that Linux could be made profitable, but they do contribute a lot of upstream code into the inner workings of Linux and it's many toolkits and daemons, but the problem is, too many of these up-and-coming developers with no notion as to what is realistic towards the goal of creating a direct alternative to UNIX are scoffing at the UNIX philosophy trying to have their own 15 minutes of fame with software that isn't worth shit.

Big name projects don't often mean they are right. Slackware is the oldest surviving Linux distribution because it does things the right way, stably, and without all the junk, garbage, and bloat as others have. You also, don't get to be the most respected Linux distribution and have a shit-for-brained software developer take a pot-shot get off without some backlash from the community and the software community at large and attacking FreeBSD blatantly without remorse or regards to what all the BSD systems have contributed to open source and even Linux itself. If anything, the Linux kernel itself owes it's existence to 386/BSD and a fatefully timed lawsuit, and owes a lot to GNU and HURD.

Now do you want to act a little less condescending and trollish, or is this going to end up like the other systemd topic in which someone got banned for shooting their mouth off without considering the consequences? Again, in this topic and what I know full well about it, I have nothing to lose.

1) No, sysvinit scripts are compatible with systemd. Read the man pages
http://www.freedesktop.org/software/...d.service.html

I'll even copy paste the relevant part for you.
Quote:
If a service is requested under a certain name but no unit configuration file is found, systemd looks for a SysV init script by the same name (with the .service suffix removed) and dynamically creates a service unit from that script. This is useful for compatibility with SysV
I won't go any further, I've done my homework, you should start doing yours before arguing with total bullshit.

I think this lecture: http://0pointer.de/blog/projects/the-biggest-myths.html would really help you to understand better. I don't want to be rude, but i'm kinda tired of these reaction.

T3Slider: I'm really glad to see someone like you. It's your choice really to don't like systemd, but at least you aren't spreading FUD like this dude.

Last edited by elvis4526; 05-07-2013 at 12:43 AM.
 
Old 05-07-2013, 12:56 AM   #34
unSpawn
Moderator
 
Registered: May 2001
Posts: 29,415
Blog Entries: 55

Rep: Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600
Quote:
Originally Posted by ReaperX7 View Post
(..) Red Hat honestly needs to just die. (..) Red Hat is like a festering cancer Linux has had for so long.
While anyone may voice their opinion this discussion should remain factual.
Please refrain from making deliberately provocative statements like that.
 
1 members found this post helpful.
Old 05-07-2013, 12:58 AM   #35
unSpawn
Moderator
 
Registered: May 2001
Posts: 29,415
Blog Entries: 55

Rep: Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600
Quote:
Originally Posted by elvis4526 View Post
If you will never use systemd because you are too fucking blind, just stop crying (..)
Personal attacks are NOT tolerated on LQ. Do not do that again.
 
4 members found this post helpful.
Old 05-07-2013, 01:01 AM   #36
elvis4526
Member
 
Registered: Aug 2011
Posts: 114

Rep: Reputation: Disabled
Quote:
Originally Posted by unSpawn View Post
Personal attacks are NOT tolerated on LQ. Do not do that again.
Personal attacks are authorized as long as it is against Lennart and Red Hat, right ?
 
Old 05-07-2013, 01:03 AM   #37
unSpawn
Moderator
 
Registered: May 2001
Posts: 29,415
Blog Entries: 55

Rep: Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600
All: and yes, that goes for all of you. Keep it factual, fun and friendly for all of us.
//This isn't a suggestion BTW.
 
Old 05-07-2013, 01:49 AM   #38
unSpawn
Moderator
 
Registered: May 2001
Posts: 29,415
Blog Entries: 55

Rep: Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600Reputation: 3600
Quote:
Originally Posted by elvis4526 View Post
Personal attacks are authorized as long as it is against Lennart and Red Hat, right ?
First of all I believe I have addressed that in my moderation of ReaperX7s post.
LQ members should have the common sense and decency to realize that attacking those who are not able to defend themselves here is not the right thing to do.


Quote:
Originally Posted by elvis4526 View Post
Personal attacks are authorized as long as it is against Lennart and Red Hat, right ?
The responsible and mature response I expect here is something along the lines of "OK, I won't do that again".
Drawing attention away from what I'm addressing by pointing at others, while itself a valid argument, does not make sense.
Do you understand that?
 
1 members found this post helpful.
Old 05-07-2013, 07:35 AM   #39
jens
Senior Member
 
Registered: May 2004
Location: Belgium
Distribution: Debian, Slackware, Fedora
Posts: 1,463

Rep: Reputation: 299Reputation: 299Reputation: 299
Quote:
Originally Posted by ReaperX7 View Post

<snip>

Systemd might be revolutionary, but it's far from being a perfect replacement to the bsdinit and sysvinit startup script kits. If the only real fix it offers is a faster boot time, then realistically for a server environment, it's completely useless to a system administrator as scripts really only require a basic text editor like Nano or VIM to edit and boot times for servers are negligible. SysVInit and BSDInit can be scripted to work in parallel the same as systemd also.

It's basically reinventing the wheel and not making it any rounder.

Supporting cgroups ... (as systemd does)
That alone is a *huge* fix (and utterly broken in all others).

PS: Solaris did the same thing using zones.

Last edited by jens; 05-07-2013 at 07:48 AM.
 
Old 05-07-2013, 08:10 AM   #40
fewt
Fuduntu Team
 
Registered: Dec 2011
Location: Nashville, TN
Distribution: Fuduntu
Posts: 75

Rep: Reputation: Disabled
Quote:
Originally Posted by jens View Post
Supporting cgroups ... (as systemd does)
That alone is a *huge* fix (and utterly broken in all others).

PS: Solaris did the same thing using zones.
You don't need an entirely new init subsystem to support cgroups, it would have been relatively simple to implement using sysvinit. The only real benefit to systemd is process management (watching a service state and taking action). Using Lennart logic, sysvinit was so broken that we should all go back to RedHat and demand our money back for those past service contracts because - per Lennart, until this new systemd soup arrived RHEL was completely worthless meaning we were being sold contracts under false pretenses.

By the way, Greg Kroah-Hartman (SUSE) and Kay Sievers (RedHat) are responsible for creating the udev abomination, if I remember correctly Greg moved on to other things and left it to be maintained by RedHat / Fedora. That's why they were so easily able to merge it with systemd.

"What kind of insane udev maintainership do we have? And can we fix it?

Greg, I think you need to step up here too. You were the one who let
udev go. If the new maintainers are causing problems, they need to be
fixed some way." - Linus Torvalds (https://lkml.org/lkml/2012/10/2/303)

You'll find a lot of "myth busting" on Lennart's site, but of course there is the word of a "Two-faced lying weasel" (Torvalds) and then there is reality.

Last edited by fewt; 05-07-2013 at 08:20 AM.
 
2 members found this post helpful.
Old 05-07-2013, 08:46 AM   #41
jens
Senior Member
 
Registered: May 2004
Location: Belgium
Distribution: Debian, Slackware, Fedora
Posts: 1,463

Rep: Reputation: 299Reputation: 299Reputation: 299
Quote:
Originally Posted by fewt View Post
You don't need an entirely new init subsystem to support cgroups, it would have been relatively simple to implement using sysvinit.
There have been many threads on the kernel list about this ...
It's pure chaos and _not_ so easy to implement at all (IMHO).

This single issue already exists over 40 years in "real" UNIX.

Honestly, most (not all) of these anti-systemd claims appear to be fuelled by an anti-Lennart/Red Hat feeling and not based on technical facts.
 
Old 05-07-2013, 08:51 AM   #42
fewt
Fuduntu Team
 
Registered: Dec 2011
Location: Nashville, TN
Distribution: Fuduntu
Posts: 75

Rep: Reputation: Disabled
Quote:
Originally Posted by jens View Post
There have been many threads on the kernel list about this ...
It's pure chaos and _not_ so easy to implement at all (IMHO).
All far simpler than completely engineering a brand new subsystem..

Quote:
Originally Posted by jens View Post
This single issue already exists over 40 years in "real" UNIX.
If it existed for 40 years, it's hard to argue that it is an issue at all.

Quote:
Originally Posted by jens View Post
Honestly, most (not all) of these anti-systemd claims appear to be fuelled by an anti-Lennart/Red Hat feeling and not based on technical facts.
Some are, but many are not. I've personally witnessed Lennart's two faced behavior first hand, but as I've said elsewhere my only remaining issues with systemd are how it was pushed into the world.

It is an example of the stigma that Linux users are all perpetual beta testers being reality.
 
1 members found this post helpful.
Old 05-07-2013, 09:36 AM   #43
jens
Senior Member
 
Registered: May 2004
Location: Belgium
Distribution: Debian, Slackware, Fedora
Posts: 1,463

Rep: Reputation: 299Reputation: 299Reputation: 299
Quote:
Originally Posted by fewt View Post
All far simpler than completely engineering a brand new subsystem..
Only if it works a 100% (and not just in some cases, while creating more complexity and breakage on the user side).
You can't simple ignore how systemd fixed this (even if you dislike everything else about it).

Quote:
Originally Posted by fewt View Post
If it existed for 40 years, it's hard to argue that it is an issue at all.
... and a bike existed way before a car, yet most people use cars.
That's not a valid argument.

As already said, Solaris did the same thing before systemd.

Quote:
Originally Posted by fewt View Post
Some are, but many are not. I've personally witnessed Lennart's two faced behavior first hand, but as I've said elsewhere my only remaining issues with systemd are how it was pushed into the world.

It is an example of the stigma that Linux users are all perpetual beta testers being reality.
I'm absolutely not saying systemd is perfect, but it's certainly not as evil/dangerous/intrusive as some people make it sound.
 
Old 05-07-2013, 09:43 AM   #44
fewt
Fuduntu Team
 
Registered: Dec 2011
Location: Nashville, TN
Distribution: Fuduntu
Posts: 75

Rep: Reputation: Disabled
Quote:
Originally Posted by jens View Post
Only if it works a 100% (and not just in some cases, while creating more complexity and breakage on the user side).
You can't simple ignore how systemd fixed this (even if you dislike everything else about it).
I didn't ignore how systemd fixed it, my argument is that we didn't need systemd to fix it. You've argued around my point here.

Quote:
Originally Posted by jens View Post
... and a bike existed way before a car, yet most people use cars.
That's not a valid argument.
Again, arguing around my point. You have not proven the issue to be an issue at all by relating it to bikes existing before cars. You have to prove an issue is an issue before you can claim something solves it. A car is better than a bike, but the sky is also blue...

Quote:
Originally Posted by jens View Post
As already said, Solaris did the same thing before systemd.
Where does Solaris stand in the market today?

Right. Solaris doing something doesn't make it right or wrong, you could also argue against it using AIX or HPUX as an example.

Quote:
Originally Posted by jens View Post
I'm absolutely not saying systemd is perfect, but it's certainly not as evil/dangerous/intrusive as some people make it sound.
Evil? No. Dangerous? Potentially to a degree if only due to the damage to the relationship between Linux and BSD. Intrusive? Yes, it absolutely is.

Keep in mind that I don't agree with Dietrich's anti-systemd articles, as they are nothing more than a FUD campaign in fact I spoke out against his use of my comment in his FUD campaign but Dietrich deleted the comment.

Last edited by fewt; 05-07-2013 at 09:53 AM.
 
Old 05-07-2013, 11:47 AM   #45
jens
Senior Member
 
Registered: May 2004
Location: Belgium
Distribution: Debian, Slackware, Fedora
Posts: 1,463

Rep: Reputation: 299Reputation: 299Reputation: 299
Quote:
Originally Posted by fewt View Post
my argument is that we didn't need systemd to fix it. You've argued around my point here.
No I didn't.
Don't use my poor English skills against me (that's not fair from a technical pov).

I'd be very happy to see a singe (non-systemd) real-life case where the kernel's cgroups are used directly without using insane hacks.

That hackery user-space fedora example mostly shows how bad this issue really is and doesn't offer a real-and-always-working fix (considering your background, I do assume you understand how bad chaotic process termination can be if anything goes wrong).


Quote:
Originally Posted by fewt View Post
Keep in mind that I don't agree with Dietrich anti-systemd articles, as they are nothing more than a FUD campaign in fact I spoke out against his use of my comment in his FUD campaign but Dietrich deleted the comment.
Glad to hear that.
Dietrich is certainly helping with 9/11-like conspiracy theories.

Last edited by jens; 05-07-2013 at 12:18 PM. Reason: removed late edits
 
  


Closed Thread

Tags
cgroups


Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is Off
HTML code is Off



Similar Threads
Thread Thread Starter Forum Replies Last Post
LXer: Slackware: Is Systemd Inevitable? LXer Syndicated Linux News 5 07-22-2013 04:54 AM
[SOLVED] slackware and systemd fl0 Slackware 512 08-29-2012 11:07 AM
slackware and systemd (OT) eloi Slackware 44 08-24-2012 04:36 PM
[SOLVED] systemd and Slackware's future kikinovak Slackware 95 07-14-2012 11:40 AM
Boot Delay 30min: systemd-analyze blame systemd-tmpfiles-setup.service BGHolmes Fedora 0 07-27-2011 09:02 AM

LinuxQuestions.org > Forums > Linux Forums > Linux - Distributions > Slackware

All times are GMT -5. The time now is 08:59 AM.

Main Menu
Advertisement
My LQ
Write for LQ
LinuxQuestions.org is looking for people interested in writing Editorials, Articles, Reviews, and more. If you'd like to contribute content, let us know.
Main Menu
Syndicate
RSS1  Latest Threads
RSS1  LQ News
Twitter: @linuxquestions
Open Source Consulting | Domain Registration