Systemd 242 Released With XBOOTLDR Support, Networkd L2TP Tunnels

[aht0]

How is my post related to this.
No they are not (i.e. Debian has no real way to do the overlay/tmpfs without custom hacking), but you surely need to learn how to use new stuff if you want to use it.
If you don't want to learn new stuff, keep using "traditional solutions", but they are not "better".

When you are 16, working with it professionally or fresh user, "learning new stuff" every few months is okay. Youth simply is more pre-attuned to trying out new ideas, professionals need to learn that stuff anyway and fresh users are "blank sheet".

Once you get to middle age and have some legacy experience then "traditional solutions" become much more appealing. "Cool new stuff" has nasty habit to break occassionally,
When you go with legacy options+little custom hacking - it works effort-wise out much the same sans frustration as going with "cool new stuff" and fixing that when it breaks.. sometimes over each update..

fixing/debugging issues in systemd, written in C 1MLoC spaghetty code..

Did he mean "Devuan"?

[Guest]

When you are 16, working with it professionally or fresh user, "learning new stuff" every few months is okay. Youth simply is more pre-attuned to trying out new ideas, professionals need to learn that stuff anyway and fresh users are "blank sheet".

Once you get to middle age and have some legacy experience then "traditional solutions" become much more appealing. "Cool new stuff" has nasty habit to break occassionally,
When you go with legacy options+little custom hacking - it works effort-wise out much the same sans frustration as going with "cool new stuff" and fixing that when it breaks.. sometimes over each update..

fixing/debugging issues in systemd, written in C 1MLoC spaghetty code..

Did he mean "Devuan"?

Systemd is been developing since years. After this amount of years, I'd aspect it is bug free, because it is the first program the kernel launches. First off, an init system should only exec the relevant services. Systemd can launch them simultaneously compared to the traditional init systems, shortening the boot time. Everything else becomes irrelevant. If it is not able to do this simple task, and only this, it's better to trash it. Then, I don't know for what reason udev has been integrated into systemd. Luckily, also udev has systemd-free alternative.
I want to say the core of each distro should be bomb-proof as much as possible. For me the core is the kernel, the init system and the necessary programs needed by Xorg (and TWM). Everything else is secondary.
Another program from the same author we use daily is buggy, after year and year of development. Maybe the priority of this person is to produce "software", regardless of its quality. Trust me when I say traditional solutions are still the best.

[starshipeleven]

When you are 16, working with it professionally or fresh user, "learning new stuff" every few months is okay.

This is completely tangential.
I'm not saying he should learn systemd

I'm saying that his statement that "traditional methods are best" is bullshit.

fixing/debugging issues in systemd, written in C 1MLoC spaghetty code..

As usual, you must add your own anti-systemd bullshit.
systemd project splits stuff around so even if there are issues you are debugging only a very small subset of 1MLoC

you have no proof to back the "spaghetti code" statement and you should feel bad about yourself.

Did he mean "Devuan"?

He meant trolololololol, that's what.

[starshipeleven]

Systemd is been developing since years. After this amount of years, I'd aspect it is bug free, because it is the first program the kernel launches.

Linux kernel is been developing since years. After this amount of years, I'd expect it is bug free, because it is the first program that the OS executes on boot.

First off, an init system should only exec the relevant services. Systemd can launch them simultaneously compared to the traditional init systems, shortening the boot time. Everything else becomes irrelevant. If it is not able to do this simple task, and only this, it's better to trash it.

So you are trashing everything that isn't SysV basically.

Are you aware of the fact that there are other inits like Runit or OpenRC or even Procd from OpenWrt that are more advanced that basic shit SysV?

Another program from the same author we use daily is buggy, after year and year of development.

Pulse has been fine for years, quit posting bs.

Maybe the priority of this person is to produce "software", regardless of its quality.

he is not alone in the team and is not the top contributor anymore

Trust me when I say traditional solutions are still the best.

Trust me when I say that what you say is the same old tired bs.

[aht0]

This is completely tangential.
I'm not saying he should learn systemd

I'm saying that his statement that "traditional methods are best" is bullshit.

It's highly subjective. For you his statement feels bullshit, for him it's plain truth. Both of you are correct in a way.

As usual, you must add your own anti-systemd bullshit.

Be weird if I, considering the thread we are in, talked about something else.

systemd project splits stuff around so even if there are issues you are debugging only a very small subset of 1MLoC

Debugging even before digging into source is tricky enough, due multitude of flags/options/settings. First you 'll have to figure out what precisely to debug and how. Erroneous behaviour could be caused by one single setting somewhere you've overlooked.

You have no proof to back the "spaghetti code" statement and you should feel bad about yourself.

Take a look at code in question? What more proof do you need? While you can to a degree get rid of modules, bunch are mandatory. It's not truly standalone per module.

"Spaghetty" is a good descriptive term for inter-locked/inter-dependent mass of code. Like intertwined individual strands of pasta forming single entity on your plate.
For ~95+% of users, following that code ain 't possible. All they can do is file bug report and hope for fix instead of "WONTFIX".

He meant trolololololol, that's what.

Yeah, whatever..

[Guest]

Linux kernel is been developing since years. After this amount of years, I'd expect it is bug free, because it is the first program that the OS executes on boot.

So you are trashing everything that isn't SysV basically.

Are you aware of the fact that there are other inits like Runit or OpenRC or even Procd from OpenWrt that are more advanced that basic shit SysV?

Pulse has been fine for years, quit posting bs.

he is not alone in the team and is not the top contributor anymore

Trust me when I say that what you say is the same old tired bs.

Well, it's quite clear you are a systemd fan. I'm only a Linux fan. This is a big difference, an insurmountable difference.
Just to say more words: the Linux kernel was, and it is still is, a miracle. Maybe you need the kerneld?