> My prediction: This will be adopted by RH first, then Debian will follow. Hopefully, Linus will stay neutral, allowing distros like Slackware to continue working without having to conform to this garbage.

i wouldn't bet a blade of grass against it. but my political posts i must cut short.

As is tradition with these types of features, you have a decent idea that gets hijacked and loaded up with additional BS for corporate control and monopolization. The idea of merging the kernel image and initrd is pretty interesting and has nothing to do with security. It simply reduces the complexity of maintaining kernel packages which can indirectly improve security (by eliminating points of failure).

But of course there are also asshats who see this as an opportunity to go way overboard and include cryptographic keys and shit in the name of "improved security". This inevitably rubs someone the wrong way, which in turn rubs someone else the wrong way.. and next thing you know some doofus is flaming everyone in broken english.

The phoronix post provides a little more context. Maybe if he-who-shall-not-be-named bothered to read all the news on this feature he might understand exactly WHY systemd is involved here.
https://www.phoronix.com/news/Fedora...-Kernel-Part-1

For the Slackware specific stuff, if this really is about merging the vmlinuz/initrd together as a single binary then that could resolve a lot of the problems people have with forgetting to run geninitrd after a kernel upgrade. But the additional benefits are so miniscule I'm not really sure I care. If someone wants to have this discussion in a Slackware specific way maybe we can take a look at what would need to be changed in Slackware to implement this. All in the name of science, of course.

1 members found this post helpful.

The inclusion of initrd in the kernel represents a security element for operation under SecureBoot, because this locally generated initrd is usually unsigned, so it is prone to attacks.

As long as Slackware is not interested in running under SecureBoot, of course these cryptographic keys are not necessary. But as you already know, the other big distributions already have support for SecureBoot. It's nothing new.

Don't you think it's interesting that even a "doofus" manages to understand that this invention of RedHat has nothing to do with Slackware?

What is the meaning of these discussions in Slackware's official forum?

Of course, distributions that use systemd will put it in initrd. For the same reason Slackware puts eudev in initrd. It's nothing new.

It seems that the only new element is this idea of universal initrd.

I have seen other discussions about this in other places and if we eliminate these hate festivals, this novelty is difficult to construct, given the conditions in which it has to solve various situations with options in the kernel command line.

If you ask me, there are few chances to see a universal initrd used in Slackware, for the simple reason that here the initrd support is treated somewhat superficially. However, it would probably be easiest to be used in Slackware, because it has an initrd that is not very configurable or modifiable.

1 members found this post helpful.

This is not true.

All ways to boot a Linux kernel work with an included initrd. Not only GRUB2 but also LILO, elilo, syslinux, refind and even direct booting from EFI or BIOS.

Because the CONFIG_INITRAMFS_SOURCE option exists starting with kernel 2.6.14 released on October 27, 2005 . It's nothing new.

https://cateee.net/lkddb/web-lkddb/I...FS_SOURCE.html

1 members found this post helpful.

You are right that this is not a Slackware specific discussion, but you fail to understand that many Slackware users are big into Web of Trust concepts.

What you, and so many other people fail to understand is that software is not deemed good/bad simply on technical merit alone. If anyone bothered to pay attention, you would have noticed that the big push for this feature came largely from the systemd community (as reported by phoronix). Does this mean that the feature itself is systemd specific? No.

But Trusting the software we use isn't about what the software does, it's about the people who write the software and their history of abuse and failure. I've seen far more "hate" from the systemd community than any other group and you clearly don't recognize this. Rejecting systemd in this way is part of the process of making sure the new feature is not specific to systemd.

When you have a Trust issue like this people tend to split into tribal groups like chimpanzees. One group will complain about Systemd, and the other group will resort to over-the-top accusations of hate and violence. Nobody hates systemd and nobody has ever hated systemd. But when Poettering says something is a good idea people throw up a big middle finger. It's a vicious circle.

If you truly want to solve these "hate festivals" then you have to confront the fact that the people in charge of systemd have said a lot of hateful things. I'm not asking you to agree, I'm asking you to understand the root cause of these types of threads. Or you can fight me, that works too!

2 members found this post helpful.

I beg your pardon?
I don't particularly care what RH or Debian do, as long as there remain functional alternative ways of doing things.

The single biggest problem with systemd is the way they're trying to force everyone to use it by making everything depend upon it.

I am very grateful to Patrick for managing to completely avoid the train wreck and continuing to do things in a sane and stable manner.

2 members found this post helpful.

Maybe that is supposed to make Linux kernel bootable in systems with Microsoft Pluton (or however this crap is called)? As there were so many horror stories of this chip making it impossible to boot anything but Windows. So maybe signed, unified kernel images are also included there. Anyway, the fact that Microsoft wants to basically control how you boot your system is just bullshit.

Anyway, I don't really see how an universal initrd could be shipped as a package, as different systems / setups might need other modules loaded (drivers, luks, etc), that's it's built on the system itself. I guess they would just be making UKI equivalent of Slackware's kernel huge or numerous variants of UKI packages.

2 members found this post helpful.

What you say, it's like the Russians to say: let's do not build airplanes because they have been invented by Americans (remembers the Wright Brothers someone?) or the Americans to say: let's do not build rockets because they was invented by German Nazis.

Or they to say: let's do not use 5G because it was invented by Chinese - oh, wait! This one really happened...

Anyway, if you look about my past ramblings about the initrd support on Slackware, you will discover that several years ago I talked about Slackware shipping an "initial generic initrd" which further could be tuned by users. Kinda like Slackware ARM do, but they repack the shipped initrd to tune it.

Yep, my friend! In fact seems like I am the one who "invented" the obvious, this "universal initrd" years ago and that Mr. Poettering "borrowed" this idea is a proof that he read even the Slackware forum.

Now you feel better knowing that those "universal initrds" was in fact invented by a Slackware user and not by them?

Probably not, because all those threads where some people expresses their gratitude for the amazing inventions of Mr. Poettering have solely, euphemistically saying... a political purpose. And of course, everybody who does not agree with those particular political views is a doofus.

1 members found this post helpful.

But you are just staring to their kitchen at what they cook for dinner initrd...

As someone will force the RedHat's initrds on you.

Yeah, I've heard this phrase for thousands times in the last 12 years or so.

I for one, as someone who does not care about systemd, I wonder WHY in all those 12 years, no programmer bothered to write from scratch system services like those offered by systemd, to escape the software World from systemd "tyranny" ?

Heck, even the EUDEV and ELOGIND are just pieces of systemd, nothing more.

No offense intended, BUT seems like that exactly the ones capable to truly and trustfully evaluate the systemd "tyranny" have a hugely different opinion.

Yeah, I've heard things like this also about LinuxPAM, about Xorg and even about UDEV...

Each time he was last one adopting those technologies, but they was adopted in the end.

I for one, I do not think that him is a systemd hater, but a very conservatory guy.

This is a very big difference. And this means that's just a matter of luck that your views right now corresponds with what Slackware it is.

1 members found this post helpful.

Well, I think the idea is to put all disk, controllers, filesystems and other associated kernel modules in initrd then to load them with UDEV (no matter that it's provided by systemd or EUDEV) and hard codding the abilities of initrd - i.e LVM/RAID, encryption, etc.

The result probably would be bigger than a huge kernel from Slackware, but the advantages would be that drivers does not step on each other shoes and the memory occupied by the drivers no loaded will be freed later.

So, I believe that an "universal initrd" combined with a generic kernel would be at least a superior solution than the huge kernel on Slackware, but honestly I have no hopes that it will adopted by Slackware. We'll see what Fedora will do.

1 members found this post helpful.

Dumbass statements like that are what make people not want to trust you. I don't know what your motive here is, but I'm going to assume malice. I trust your contributions to Slackware to the extent that they get filtered by Pat and the rest of the team.

Tell your propagandist that Robert H. Goddard says hello.

2 members found this post helpful.

Well, the first functional ballistic missile was built by Nazi Germany, Aggregat 4 was known as the V2 (Vergeltungswaffe 2) and was used for the bombing of London.

https://en.wikipedia.org/wiki/V-2_rocket

After the war, one of the creators of V2, Wernher von Braun and a team of German scientists laid the foundations of American ballistic missiles and the rockets used by civil agency NASA, culminating in the Apollo missions, which took men to the Moon.

https://en.wikipedia.org/wiki/Wernher_von_Braun

At the same time, another group of 170 German scientists, led by Helmut Gröttrup, laid the foundations for Soviet ballistic missiles and the Soviet space program, culminating in the launch of Sputnik 1.

https://en.wikipedia.org/wiki/Helmut_Gr%C3%B6ttrup

https://en.wikipedia.org/wiki/Sputnik_1

As an interesting note, after withdrawing from the Soviet space program and returning to Germany, Helmut Gröttrup invents the credit card for Standard Elektrik Lorenz.

History says that all these German scientists, working after the war for America and the Soviet Union, were former members of the German National Socialist Party. However, they are the same people who built rockets that carried people into space and gave the great powers intercontinental ballistic missiles.

But I'm sure LuckyCyborg only used that phrase to ironically point out the absurdity of rejecting a technology just because it was invented by a particular engineering team.

1 members found this post helpful.

The scientists from former Nazi Germany also built (after WW2) for Soviet Union both nuclear plants and nuclear bombs. Historians believes that the Soviets stolen (via espionage) the nuclear technology from USA. That's not true.

The Soviet nuclear bombs was built by the German scientists who was "invited" after war to continue their work in USSR.

Yeah, the Nazi Germany had also a nuclear program, but fortunately for us, they was initially focused in producing energy (nuclear plants) and only later got idea of making nuclear bombs. Too late for and when the resources was scarce because the evolution of war.

https://en.wikipedia.org/wiki/German...eapons_program

Apparently, on top of their list of targets for nuking was Moscow. Yet, Joseph Stalin and all Soviet leadership does not given too much attention to nuclear devices until the bombings on Japan. This is moment when USSR wake up and started working on nuclear bombs using the German scientists borrowed from the Soviet occupied territory.

Thanks for understanding my point.

Believe or not, I have no ulterior motives?

Or you believe that I am a mole planted here in 2010 with the purpose to steal the Slackware technology?

After all, everything could be downloaded gratis, as you know. No espionage is required.

Thanks you! I find my fun on improving Slackware and I am glad when some of my suggestions are accepted.

I know about Robert H. Goddard, also I know about Hermann Oberth, the Austro-Hungarian (later German) scientist and professor who worked with his former student Wernher von Braun on building the V2 rockets. And Saturn V.

Also, as a trivia Hermann Oberth published in 1923 his doctorate book on mathematics of spaceflight with: Die Rakete zu den Planetenräumen , aka "The Rocket into Planetary Space" , which is considered (at least on some place of the World) as being the theoretical basis of functioning of space rockets (and ballistic missiles ).

If I will ask a scientist in my country about who invented the rockets (as mathematics), he will say Konstantin Tsiolkovsky, and yes, he did this in 1903, but his equations was improved by Robert H. Goddard in 1912, and finally Hermann Oberth got a complete solution of those mathematics in 1920. The history says that they are the Founder Fathers of the rocket science.

If that is correct, then he doesn't really understand what's going on here.

The reasons for this technology seem to be quite sketchy, but the major players are eating it up.
No, but the situation is significantly worse than that.
Maybe because it's simply not required? Linux had already achieved global domination on servers before systemd came along, and it got there without service supervision.

Then someone came along and said, "Hey, your T-34 doesn't have any spare tyres!"
You might be rewriting history there. I don't remember any such discussions about Xorg or Udev.

By delaying the adoption of PAM, Patrick avoided decades of problems. PAM was a complete train wreck for many years. The result we have now, with PAM and Kerberos in Slackware 15 means that you can drop it in as a Samba DC and it works flawlessly and completely invisibly to Windows users. PAM was adopted in a sane and stable manner. As far as I'm aware, there haven't been any security patches for 15.0 related to PAM so far.

2 members found this post helpful.