Red Hat's contributions are many but often not everything they add in upstream makes it way into the stable kernel. Gentoo, Ubuntu, and other big brand name Linux distributions contribute the same amount into the kernel and all their stuff ends up as custom patches to the kernel, not mainstream.

The most important part of the GNU/Linux system Mercury isn't the kernel, but it plays a large part in it, but the kernel is NOT the most important part. The most important part is the Shell Environment that interfaces systems tools with the kernel. To be frank, BASH (and any other shell environment) is the most important part of the entire GNU operating system. Without it, nothing can function, and no work can be done.

BSD is a kernel also, but each distribution has its own tools. FreeBSD is FreeBSD through and through. Same with every other BSD, or BSD-based distribution. Each has tools specific to their own operating system. The only difference is PC-BSD which is a prebuilt FreeBSD based distribution. BSD has the right to laugh at Linux because they can and have every right and reason to.

Linux is NOT largely created by Red Hat. Linux based systems are largely created by the GNU Foundation with contributions from FreeDesktops, The Open Group, The Linux Foundation, and hundreds of thousands of individual developers from all over the world not even affiliated with any organization at all.

Red Hat's role has been significant in MARKETING LINUX, but minor by comparison to the nature of the whole in terms of actual developments across the spread spectrum.

Just because Red Hat says to the community "Jump!" doesn't mean everyone in the Linux development community has to reply back, "How High?". In fact many probably would and do reply back, "Piss off!"

I don't believe this. I didnt even search about it. Send me a reliable link.

et tu Brutus?
, yes like being the most important part in an OS?, Is this a debate? Lets go back open the dictionary and read definition of Important. NoHmm.. Android?

Correction: FreeBSD has a Kernel FreeBSD is a full OS that includes shell userland. Agreed

Linux is a Kernel not an OS. RH contributes most to the Kernel hence Linux. That is called the GNU userland.

FALSE

Thats the problem they don't force anybody. They put up their standards and give you the freedom to use them or not. Why else would Slackware not be using RPM, Yum, systemd, pulseaudio and all the other RHEL tools? Answer: Freedom.
There will always be people that blame others for their problems. This is just life. But one needs to at times point the finger at themselves and say its my fault not RHEL. Blaming other distros is not the correct mind frame to work with imho.

So let me guess what you are thinking now? Systemd is not cross platform.

OK... Why should it be? Why should Lennart design cross platform. Heck, its his creation. He can do whatever he wants with it. Put yourself in his shoes. You spend all this time working creating something for your distro. Then you got all these FreeBSD peoples crying saying it should be Cross Platform... Heck... Why? Its your creation, you can code to do whatever you want? What is this some communist manifesto? You have to design cross platform for all Unixes? Last time I checked FreeBSD hated Linux guts... and rightfully so. When has FreeBSD contributed to Linux or designed things cross platform?

Ohh but they are Unix? and Unix Philosophy. Give me a break Reaper. I respect your belief in Unix Principles but you gotta respect a Developers Creation as well. Let the man code for himself and his own needs or desires. Even if it is evil let him do his thing don't hate. In the end you are just a user. If you don't like it code yourself or use the old sys v init. Why hold him back from his projects? If systemd is as bad as it is then nobody will support it and it will as Eric likes to put it Kaput!

Red Hat has contributed much to the GNU/Linux community and served as one of the first proofs that it is possible for open source projects to make money. However, Red Hat does not 'own' all of the projects that they contribute to, and they benefit from the work of others. Thus, giving Red Hat the exclusive ability to decide the future of GNU/Linux based on their contributions would be erroneous. According to this link Red Hat is the highest-ranked single commercial contributor to the kernel -- however, the number of non-commercial or commercial unaffiliated developers both (separately and together) out-rank Red Hat. Red Hat definitely deserves some credit, but so do Novell/SUSE, Intel, IBM, etc. With 11.9% of the kernel code contributions, Red Hat has no business being the sole decider of Linux's future. It is more difficult to find statistics on low-level userland code contributions since the various projects are diverse and many, and I would expect Red Hat to throw more weight behind that -- however, to give Red Hat carte blanche to do whatever they want when every major *nix userland utility receives significant contributions from diverse sources is ridiculous.

This of course has little to do with the discussion and I don't understand why the discussion turns to distro wars.

Agreed. Agreed. I never said they contributed more then "everyone else combined". Agreed. Disagreed. They can do whatever they want. Just as Slackware or any other distro can do whatever they want in their own linux distro.

Agreed. Its the nature of Slackware Forums to turn things into discussions. Long before I even became an LQ member. Maybe thats what attracts some people here?

Okay, let me go slightly off-topic here a second, and I want this to be public.

Mercury don't you EVER pull a "troll-bash" (line by line attempt at discrediting) against me. I've been more than respectful and professional with you till now. I've tolerated you in PM and open forum and even held respectful chats with you. If you want me to start ignoring you and think less of a person that you are and can be, don't "troll-bash" me ever again. Unless you ARE a troll who likes hearing themselves talk, has to have the last word, picks apart statements just to get a rise out of someone on purpose, and has to be correct regardless if they are wrong, please be respectful with me as I have with you. This a professional forum with people using respectful levels of chat between themselves and with others, not a filth ridden backwater forum where anything goes.

Back on topic...

Exactly, the amount of upstream code is plentiful, but almost none of it in practical in a vanilla sense. Most of it is vendor only specific patches you count the Ubuntu, Gentoo, Red Hat, etc. patches and you get plenty of contributions, but none of them are general usage and practice practical. You said it correctly T3Slider:

That's more than enough weight to say Red Hat is NOT a majority developer. Commercially they are, but non-commercially and on the whole, they're small potatoes, and even Novell/SuSE is a major UNIX maintainer and owner of UNIX code and copyrights.

Sorry if I appeared to come off that way to you. I honestly did not want to disrespect you in anyway. But I did totally disagree with a lot of what you wrote there. Pretty much most of the posts were directed to my responses where they not? I would definitely not like to argue with you in anyway especially not in a forum. Hopefully this won't happen again getting you upset is the last thing I want. Honestly I am as surprised as you are for your responses towards mines. You have been cool with me since I first signed on to LQ which is why I wrote you PMs.
Anyways peace.

http://forums.gentoo.org/viewtopic-t-934678.html

udev forked, apparently by or including Gentoo user(s). Not sure how experienced the forker is but I hope they use the suggested name µdev.

So answer this question Mercury...

If bsdinit and sysvinit are simpler tools that are still functional in every bit the same way as systemd and aren't as intrusive to the system, system resource heavy, and less modular by design, how is systemd better if sysvinit and bsdinit can perform the exact same functions, lessen the load on the system, aren't intrusive and require large initramfs loads and a split /usr partition, and can be utilized by more systems than just a GNU/Linux based distribution?

My answer may vary from yours but here's what my opinion is:

Systemd is not a better init system overall and here's why.

1. It's too resource dependent which breaks compatibility factors with older hardware Linux aims at continued support for. Going for 16mb RAM to 64mb RAM is a vast change if you are a system operator.

2. BSD and SysV init systems can scripted to work parallel functions and boot systems the same as systemd without replacing the entire stack of functions and protocols.

3. Linux needs to remain modular because modular works. Modular allows for errors to occur without bringing down the entire system, a lesson hard learned by Microsoft with MsDOS, Win9x/ME, and early models of NT. A program could crash and simply be relaunched.

4. Splitting /usr creates more bulk in design of library links and such. More tools and more functions have to be added just to replace what exists in simplification with complexity. If systemd promotes the KISS principle, how is creating more and more instances of complexity a KISS principle? It's not regardless how you choose to answer.

5. Is speed all that important? Will booting your system faster make it run better, more stable, and with higher compatibility? No it won't. The battle of speed versus compatibility has always sided with compatibility.

6. Compatibility is king when it comes to software and because Linux is a GNU based UNIX-like system, having tools and software that is cross-platform compatible allows systems like Solaris, BSD, Illumos, and other kernels and systems to interact with minimal additional compatibility layers with Linux based software. BSD, Illumos, Solaris, and various other production level kernels all, on their respective operating system distributions contain mixtures of Linux based software to add functionality to their systems. This plays into the Open-UNIX Platform design completely.

"One format to greet them, one toolkit to run them, one specification to follow, and in an cross-system standard unite them." That is the Open-UNIX Philosophy, Lord of the Ring's Style.

1 members found this post helpful.

Already posted.

http://www.linuxquestions.org/questi...ev-4175424319/

Apologies, did not see it. I'm subscribed to this thread.

I respectfully would like to decline this argument with you. imho systemd is still in testing (although not stated) and has a lot of advantages along with some disadvantages. I have used both systems and have some sort of opinion about each from my use. I would recommend you to also try it out and see it for yourself to have a better opinion and understanding of it. Fedora 17 is a good way to find out. Perhaps in a kvm if you dont want a full install.
In the past when I first started using Slackware and comparing it. I had a lot of negative things to say about the distro due to my lack of knowledge. But as my knowledge increased (the more I used it) I seemed to appreciate each distro individually rather then focusing on their negative traits.
I think its wrong to believe anyones word on a forum including mines. I believe the best way to understand something is to try it out. However a forum will help give you things to focus on certain aspects. This is why I use this forum. As I stated to you in my PM. Its not about making friends in a forum that is important. It is about learning and asking questions that will give you the answers any means necessary. That is pragmatism. I asked many questions here then later tried them out and done searches on google to expand my knowledge on what I was able to grasp. I learned a lot in a short time and some of that is due to this forum. So if certain people in the forum hate my guts its ok with me... as long as I take my knowledge and advance... It is me that has ultimately won. Knowledge is the true prize here not winning an argument. I hope I lose more arguments in the future... that way I learn. I have lost a few arguments that have changed the way I think of Slackware for example. As I said its not about winning in an argument its about learning from the argument that is most important. So why am I not arguing with you about your questions. Because I want you to try yourself and learn on your own about systemd. I think you have been in here too long discussing systemd and you have not yet tried it. "Doublethink"

Mercury, I don't think you've grasped the fundamental issue here.

We don't CARE if systemd is better or not, we don't care for its features. We don't care if it can function as a bread toaster. We don't care if it will boot our systems a few seconds faster.

We don't want it because we have a growing list of preliminary objections to its design that seeks to break the simplicity of underlying UNIX-like system and break existing well-tested and stable solutions which we value over slight increase in performance.

I don't know how much simpler I can put that. If Linux starts adopting these things, I will consider BSD. Seriously.

2 members found this post helpful.

No prob. It was only posted with a title called udev, so it might have been missed.

Ironically, I'm not surprised by this reaction is starting to happen but I'm shocked it happened so soon actually.

Many distributions are starting to weigh their options, and this backlash against systemd is probably going to split Linux straight down the middle with the factions of "Systemd Linux" and "SysV/BSD-Init Linux" with each having their own tools and fundamental systems.

I won't be shocked at home many others are not just going to help out with this, but contribute code, and maybe even completely fork out µdev from udev into it's own stand-alone system toolkit.

The UNIX Philosophy has been the backbone of Linux, BSD, Solaris, Illumos, and every major and minor kernel out there along with every software package, subsystem, and distribution. Open-UNIX as a standard has brought all kinds of software not just to Linux but BSD, Illumos, Solaris, etc. Open-UNIX as a standard has defined what GNU is, and what Richard Stallman was trying to say, now that I've though about it.

It's not just about "free software"... it's about the freedom of that software to work where ever it's ported. That's what Stallman never could make clear to people. Freedom is not free when it's bound to only one thing. Linux is about freedom, GNU is about freedom, and software should be "free" to run on any platform is can be designed and ported for with minimal effort required to do so if the right tools exist to make it so.

1. This is totally false. If you compare basically sysvinit and systemd, yes systemd is a little bit heavier BUT, systemd replace consolekit and udev. If you are comparing the memory usage of both, you must count sysvinit + udev + consolekit. This is totally FUD.

2. The advantages of systemd aren't only parallelization of the boot process.

3. Systemd is in fact like 7-8 binaries (more or less) and guess what... YOU CAN ACTUALLY USE THOSE YOU WANT AND REMOVE THOSE THAT YOU DON'T WANT. If systemd crash, the ENTIRE system does not freeze. This issue was asked to Lennart a lot of time. If you would actually tried systemd for real, and read some documentation about it, you would have known.

4. You speaking about KISS principle... I hope you know that on Unix, originally, /bin was a symlink to /usr/bin ? I don't see the complexity here, it's just a symlink.

5. No, boot speed is not that important but it is nice to have a fast one. Don't speak like the entire GNU/Linux community thinks like you. Some may find the boot speed irrelevant, some may find it important. It depend on the context. On a desktop, it's nice to have a fast boot speed. On a server, personally, I don't actually give a fuck.

6. You know, it's the people that makes the software that dictate the rules. This is the way Linux is taking, because a majority of people seems to think it is the best way. If you aren't happy with this, you are free to migrate on BSD that should be more suited to your need for what I can see.

Guys, before ranting about systemd, just try it and read documentations about it. If you didn't, then you are misinformed and spreading false fact everywhere.

2 members found this post helpful.

Trinity forked from KDE and it has essentially gone nowhere (and I mean no offense to the Trinity project). We can only wait and see what happens to the udev fork. If it can maintain API compatibility with systemd's absorbed udev then all can be appeased.

The last thing I want is a franken-SysV init that spawns a bunch of instances of bash to attempt to run multiple things simultaneously. If you are going to parallelize the boot process, systemd (or a similar system without the absorbed udev) is a MUCH better way of doing it.
systemd is not all contained in one binary -- it contains many individual parts, just all wrapped into the same *project*. The chances of the system crashing because of systemd are not much higher than with SysV init, if at all. The youthfulness of the actual pid=1 process may present the opportunity for bugs to exist at this point, but in general the design of systemd does not fail because it is not modular -- it is modular.