systemd has been a complete, utter, unmitigated success
systemd has been a complete, utter, unmitigated success
blog.tjll.net systemd has been a complete, utter, unmitigated success
Eleven init systems enter, one init system leaves.
You're viewing a single thread.
134 comments
-
I don't appreciate the attitude and arrogance of the guy behind systemd because he actually believes what he produces can replace everything that already "just works". He wants to push out systemd-homed because "why not". He wants to replace grub. He wants to replace a myriad of things that just flat out don't need to get replaced. autofs, cron, you name it! That kind of thinking and one-size-fits-all mentality is backwards and does not benefit the community in any way. All it does is stuff everything into one bin and so long as influencers like this guy continue to restrict what works or doesn't work according to their own work, the community and its users will not be able to freely develop FOSS. Gnome is a good example of something that creates too much of a dependency on systemd and so when you're trying to use something like Gentoo, it becomes very difficult to get that done and hacks have to made in order to get it working. FOSS shouldn't work like that. He'll keep stripping away legit projects from major distros until IBM/Red Hat finally decide to seal the deal and lock everyone out for good. Sorry if I can't rejoice in the woah whiplash.
-
The is the first time I’ve ever heard someone accusing grub of „just working“
-
Grub is working perfectly fine.
If it breaks it is, in my experience as a grub user for over 20 years and as a guy working in server hosting for 15 years, either because of failing HDD/SSD or because of user error. People don't read when the updater tells them that running "grub-install" is needed (or they perform it on the wrong drive/partition) and then blame grub when it fails on the next boot.
The crappy bootloader that comes with systemd very often, in my experience, fails to register that a new Kernel was installed and boots the old one (or fails to boot if the package manager removed the old Kernel).
Oh and GRUB has so many useful features, like booting a ISO image. GRUB is a piece of programmer art!
-
lmao
-
-
All it does is stuff everything into one bin
Well, it is not one bin. There is no monolithic systemd bin that does everything. There are a lot of separate bin files for all the different tasks. Well and if you don't want to use timers, then don't and just use cron instead. If you don't want to use journald, then just don't and use rsyslog or whatever you want. Don't need systemd-homed? Well, then don't use it. You want to configure your network with something else then systemd-networkd? Great, do it if you want.
The Poettering Army will not come and force you to enable all the options 😜
-
Except, they are. Pottering is the front man who does the dirty work for IBM and Microsoft to take over Linux by forcing distros to adopt systemd.
Those of us old enough to remember the "vote" that resulted in Debian going to Systemd remember it was almost at gunpoint.
Death to systemd, long live FOSS culture
-
I am not seeing how IBM and/or Microsoft are winning anything here or how systemd enables them to take over Linux. But maybe I am missing something.
Last time I checked (60 seconds ago) systemd was using FOSS licences for all it's code. So it seems to be living the FOSS culture, or not?
I am always open to learn and correct my view on things under new information, so if you can provide them I am open to read it.
-
Ah but you see, you have to understand the FOSS community a little more than just "using a license that FSF and OSI endorsed".
In terms of inter-project politics, systemd is almost wholly owned by IBM. They can override any will they want, they can change anything they want, all while fucking the community over. In short, IBM, using systemd as a massive octopus growing it's tentacles all over mainstream Linux distros, is gaining considerable weight to pull in the Linux world.
They can essentially dictate matters to everyone they want, because you don't want your distro to stop being supported, do you? And now, another IBM-majority project, GNOME, is almost dependent on systemd (despite the very good word of both gnome and systemd that this wouldn't happen, IT HAS) and KDE is also being slowly pulled in that direction, with DrKonqi becoming systemd only in it's latest update.
Essentially, we are handing over 30 years of work in FOSS to IBM, literally the caricature of evil tech company, and now they control the mainstream and can dictate their will.
Allow me to remind you that this same IBM almost immediately after taking over RedHat, started closing down the source sharing of RHEL, which is it's own whole thing so I digress.
Let my final word be this, R.M.S as much of a problematic piece of shit he is, correctly predicted we being fucked over by DRM and subscription services 20 years ago and was ridiculed for it.
Don't you think it's time to take a fucking hint? You don't have to be an anarchist to see where this is going.
-
I have seen with Oracle Java and OpenOffice (as two examples) that the open source community is very good in just leaving and forking a project if the current owners fuck up.
The same will happen with systemd if needed. Red Hat may be the primary source behind systemd now, but they don't own it. All the code is fully open source, none of your ramblings have any hint of facts or any real foreseeable danger behind it. I asked for facts, for anything with some kind of real information behind it.
There is nothing that powers the claim that RedHat or IBM could take over Linux with systemd. How would they do it? They can't, because even if IBM would tomorrow change the license to a closed one and would want money. Who cares, everyone will just fork the version before the license change and good is.
Just as it happened back then with Xorg (I mean the change 15 or so years ago, not the current strange fork), like it happened a short while ago with Redis, and there are so many examples more.
-
-
-
-
-
He wants to replace grub.
I hope so!
-
So, I don't like the guy either, but for a little devil's advocacy:
The stuff that already "just works" was developed during a very different era in terms of computing power, tasking of the computers which were running the systems, etc. Nobody (serious, and he is serious) develops something different because "why not?" they, at least from their perspective, feel that they are improving on the status quo, at least for the use cases they are considering.
one-size-fits-all mentality is
being decided by the distro maintainers, not the developers. Sure, developers promote their product, but if a distro thinks that multiple flavors are a better path, they distribute multiple flavors. It's not like the systemd developers are filling billion dollar war chests with profit because they're using strong-arm tactics to coerce distro maintainers to adopt their products.
stuff everything into one bin
When one bin serves the purpose, it's a lot easier to maintain, modernize, security harden, etc. than ten bins.
the community and its users will
notalways be able to freely develop FOSS.Fork it and your loyal users will follow.
Gnome is a good example of something that creates too much of a dependency
Agreed, I was never happy with GNOME, and starting about 5 years back I have been migrating my systems, personal and professional, off of it. That's the nature of FOSS, no contracts to negotiate, make the choices that make sense for your use cases and execute them.
FOSS shouldn’t work like that.
FOSS, by its very nature, should be expected to work all the ways. If a particular way can't get enough developer traction, it stagnates but never really dies, not until the ecosystem it is dependent upon can no longer find hardware to run on and users willing to run it.
IBM/Red Hat finally decide to seal the deal and lock everyone out for good.
I am very glad that I walked away from CentOS about 8 years back, its proximity to Red Hat never made me happy. I have been trying to walk away from Canonical (toward Debian) for about 3 years now, but it still has some hooks that keep our professional team happier than Debian. If the unhappy ever outweighs the happy, we'll execute the move.
Sorry if I can’t rejoice
Never asked you to. End of devil's advocacy. I still don't like the guy, but I never really interact with him. I do interact with his products and the alternatives, and in my use cases the products speak for themselves. There's nothing about systemd that makes me dig around for systemd free alternatives - they are out there, but for my use cases I don't care. YMMV.
-
Why did you quote me but leave out where I mention systemd explicitly with Gnome? lol
So you agree Gnome has too much of a dependency on systemd. Let's not beat around the bush. Let's call a spade a spade.
Does Gnome have too much dependency on Gnome: yes or no?
-
Gnome is a good example of something that creates too much of a dependency
Agreed, I was never happy with GNOME, and starting about 5 years back I have been migrating my systems, personal and professional, off of it. That’s the nature of FOSS, no contracts to negotiate, make the choices that make sense for your use cases and execute them.
Does Gnome have too much dependency on Gnome: yes or no?
Absolutely. If you don't mind using Gnome exactly as Gnome wants you to - this year - then it's usually a pretty refined desktop experience, but if I wanted to be told what to like, how to like it, and to shut up and be happy, I'd use a Mac.
I prefer XFCE for its modularity... don't want a launcher bar? Don't run the launcher; nothing else misses it when it's gone.
Mess around with Gnome too much and it becomes a nightmare mess of dependencies.
-
-
-
Grub is shit
-
Aye lmao
-
-
134 comments