Being strict does NOT mean or imply being an a*hole. If you aren't strict, the a*holes will take over and turn everything into sh*t.

6 members found this post helpful.

Actually, everything I've ever read about Linus' more colorful statements amounts to this: He gets grumpy with people that should know better. Not noobs, not people whom are learning or don't have a clear picture. i.e. Mistakes that are made from carelessness, apathy or arrogance by people whom have the experience not too.

I can't say I blame him. I've got a guy at work who is a Sr. Admin and team manager who has said the most breathtakingly stupid things I've ever heard. I'm really, really glad I don't work for him because I could not sit there while he told me stuff like, "We shouldn't use x tool in x situation just because that's what its designed to do." So... then we should use a screwdriver to do a hammer's job? Are you mad? - No way I could keep that rebuttal from coming out of my mouth.

1 members found this post helpful.

Should I write:
this boring thread is useless
or
this useless thread is boring
?

1 members found this post helpful.

You shouldn't comment at all, then.

9 members found this post helpful.

I think it's clear that the goal now of systemd is to create an OS (I've now seen two quotes from Poettering to that effect). I suspect he (and others) sees the Linux kernel as part of that process and Linus will simply need to fall into line at some point. It's pretty arrogant.

I suppose the whole process will ultimately see a transformational change in what a large chunk of the Linux universe is, for better or worse.

So the designer of a ZF 6 speed must understand that a 265 V8 won't be able to deal with the ratios? Sorry but that's just wrong. The fault is with the owner of the vehicle who wants to fit a part that is not suitable. Either fit an appropriate gearbox or if you desperately need the 0.50 overdrive ratio in your car fit an engine as well that can handle it.

Why is it these discussions all go like this? Got nothing reasonable to say so someone posts analogies that are ludicrous.

The problem between systemd and the kernel indicates the kernel has a weakness, admittedly a weakness that would probably never have been found unless something like systemd come along but the thing is it has and it has shown this weakness. Now that the weakness has been shown to exist wouldn't it be prudent to fix it? Isn't fixing problems the Linux way? or are we just going to ignore this weakness in the kernel because it is systemd, and by extension systemd's creators, who through no fault of their own have shown it exists?

I might be wrong on this one but can't we say sytemd spammed so many logs that it was like a DDoS attack? If so, I don't think it is kernel's fault.

The systemd people, especially Poettering, come from the angle that the current state of GNU/Linux is a mess, too many choices, over reliance on shell commands, not enough standardisation and dependence on people who know how to write scripts, so their objective is clearly not just an init system.

Poettering's views have been made clear time and time again - one famous example is:

"Linux is still too fragmented" http://linuxfr.org/nodes/86687/comments/1249943

It's pretty obvious that Poettering's vision is an OS which runs on his "standardised" software with Linux as the kernel and gnome 3 as the desktop.

MacOS gets mentioned by Poettering no less than 5 times as either a benchmark or inspiration - it's clear that this is his wet dream of the ideal OS/experience.

People like Poettering and de Icaza are selling the GNU/Linux eco system down the river to meet their own agendas. They're on power trips and sadly they're having their arses wiped every step of the way by other developers who should know better.

13 members found this post helpful.

So users that are trying to fit systemd in Linux are to blame? ) You've missed the analogy.. Probably because you either work with cars or actually know a bit (a lot) more of them then us... Think of a gearbox that's made for a specific car/engine, but fails.. After just putting in the gearbox and it fails, who do you blaim..

Anyway, I strongly suggest both of you to give up this analogy .. It really is beyond the scope..


The kernel had a weakness and they fixed it.. That's exactly what the systemd guys refused to do in their part.. They are the ones that said the debug messages are bloated to the point that they are not useful.. So then why keep them in the global 'debug'? Well.. because apparentely "they are building a core OS", and the dev kernels or users that actually use that debug statement (and have done so in the last couple of years) well.. "they all suck".. It was never about anything else...

Btw, from my understanding, the bloated messages was an already fixed bug (still, either Kay did know and just behaved like a child, or did not know and behaved like an idiot.. ).. Weird, what if that would have been a regression.. Would it this time not be a bug? Anyway, even now, when they are asked to use systemd.debug instead (and this is just because Linus doesn't trust them), they still insist on using the global parameter and just switch the default target of the debug messages (which, at least, is a small step forward)..

I can't wait until the day comes to "jailbrake" my Linux system..

As a user of a "[not] really too relevant" OS / "toy OS" you don't really have much to worry about.

"It is impossible to make anything foolproof because fools are so ingenious."

this "weakness" needs root privileges
there are many more ways to break the kernel if you have root privileges

Nah, you see I never said that nor did I miss the analogy. What I did say was the analogy used was wrong.
So you're trying to re-word the analogy to make it fit. Sorry still doesn't work.
Why then get involved in it if you want us to drop it?

Smokey I haven't heaped praise on the systemd devs, I have pointed out that without systemd the problem in the kernel may not have been found and therefore the kernel would have a weak spot.

Study the configure script, you can disable most of the features (AFAIK all except journald and udev) at will.
Really, before commenting further you should use some time to investigate systemd a bit further, the lack of modularity you speak of does not really exist. You may also want to have a look at bartgymnast's effort to port systemd to Slackware, just to see how minimal a systemd installation can be.
Yes, systemd requires glibc, yes, the systemd people prefer not to code POSIX compliant, but to use all the advanced features of the Linux environment instead. This is all known for a long time already and shouldn't surprise anyone anymore.

Both parties are at fault here. Yes, systemd spammed kmsg (by the way, this was a bug that at that time was already fixed in their git), but kmsg had no mechanism in place that could handle the spamming, which at leat I consider to be a security problem and therefore a bug. You may consider that a DDoS attack, but not being able to handle that is the kernel's fault.