postmarketOS just gained my respect. To be fair there's no point in running a Linux system without systemd as you'll end up installing 32434 different RAM wasting services to handle things like cron, dns, ntp, mounts, sessions, log management etc.
Linux
From Wikipedia, the free encyclopedia
Linux is a family of open source Unix-like operating systems based on the Linux kernel, an operating system kernel first released on September 17, 1991 by Linus Torvalds. Linux is typically packaged in a Linux distribution (or distro for short).
Distributions include the Linux kernel and supporting system software and libraries, many of which are provided by the GNU Project. Many Linux distributions use the word "Linux" in their name, but the Free Software Foundation uses the name GNU/Linux to emphasize the importance of GNU software, causing some controversy.
Rules
- Posts must be relevant to operating systems running the Linux kernel. GNU/Linux or otherwise.
- No misinformation
- No NSFW content
- No hate speech, bigotry, etc
Related Communities
Community icon by Alpár-Etele Méder, licensed under CC BY 3.0
Whether you like Systemd or not, let’s not spread disinformation.
All these things still exist with Systemd. They are just called Systemd dash something. Also, while Systemd is feature rich, it is pretty heavy relative to many alternatives.
Distros that avoid Systemd typically do so because they consider it bloated and possibly insecure.
If you are a fan of Systemd, it is probably because you like the standardization and the integration across previously disparate services. That makes sense. If you think it is making your system faster or lighter, you have not explored the alternatives. Obviously Systemd was a big leap forward in init. Other systems have appeared that also work really well but they are probably too late to matter mainstream. The “market” has spoken and Systemd is the winner.
I've yet to find a use-case for "making my system lighter" by exchanging a daemon that takes <0.1% of my total system memory for a bunch of poorly maintained bash scripts.
All these things still exist with Systemd. They are just called Systemd dash something
Although you're right that's not that "cut and dry" there's a lot integrated and baked into the systemd core. Even if you consider a "systemd dash something", let's say systemd-networkd
we're suddenly talking about a single efficient daemon that covers all networking from basic IPv4 DHCP to IPv6 (with all it's possible addressing schemes), can act a client or more like a typical router acts, delegate stuff and manage the entire thing from top to bottom in a cohesive way.
Just think about the amount of crap you've to setup to have a system do dual stack networking and provide prefix delegation on another interface, with systemd it's just systemd-networkd
. From the basics to serve IP's, the classical isc-dhcp
can do both IPv4 and IPv6, however...
the ISC DHCP server can only serve IPv4 or IPv6, means you have to start the daemon twice (for IPv6 with option ”-6”) to support both protocols.
Or you'll just find you the implementation is bad and you'll run wide-dhcpv6
instead. And then you won't survive without radvd
for router advertisements etc.
If you are a fan of Systemd, it is probably because you like the standardization and the integration across previously disparate services. That makes sense. (...) Obviously Systemd was a big leap forward in init.
Exactly, systemd solves tons of painful issues and provides a cohesive ecosystem of tools to manage Linux systems. While there are other great alternatives none as are complete and solid as systemd.
If you think it is making your system faster or lighter
But it may. By not having to deal with bunch of poorly integrated tools such as dhcpcd
, dhcpv6
, radvd
, chrony
, NetworkManager
, resolvconf
, logrotate
and others we might actually have less overhead. And I'm not even talking about the time we don't have to spend making sure all those integrate properly learning 234 different configurations syntaxes and dealing with specific bugs that only happen when program X interacts with program Y with feature Z enabled.
I'm not saying system it perfect, because it isn't but it sure provides a LOT of piece of mind, stability and makes Linux a lot better than it used to be with init and friends.
The only time it does make sense is on minimal systems like routers
because we all know that routers have so much RAM that installing DNS, NTP, mounts, session, log management isn't a problem? something doesn't add up...
I believe @[email protected] 's point was that in OpenWRT and others it makes more sense to have smaller daemons instead of systemd because they aren't using the standard ones you'll usually find under Debian and other Linux distros. They take daemons and slim them down to the point they become smaller than systemd at the cost of features that aren't required on routers.
Exactly, although it applies to more systems other than OpenWRT
Routers lack storage and RAM both of which are used up by using a heavier init. Most of the time you will see a very basic system start services by putting them in init.d
Man my router has 512 Gigs and 16 gigs or RAM. R
Mine has 128mb of ram. What on earth are you running on your router than needs that much hardware. I just bought a device from Walmart
Probably running OPNSense in a VM.
With 16gb or ram?
Rams cheap. Maybe he’s getting full ipv6 routing tables.
Next step is drivers that allow switching on/off components of the phone through systemd to save battery. Proper drivers are the only major missing piece for Linux phone OS right now.
When Linux phone comes with lasting battery and fast waydroid I'll switch.
And firmware updates haha
Well that's harder
What firmware do you mean?
The firmware of the device components, that needs to be signed by the manifacturer and often has critical bugs.
I see, but which components do you mean specifically?
Modem, wifi/bluetooth chip, many more possibly.
In case of the modem I know it doesn't need to be signed. In fact, there exists an open source firmware for it.
Cool! Thanks for the info
Exciting to see! Positively surprised Alpine is modular enough to make this feasible/maintainable.
Curious to see what the part about SystemD and musl at the end meant.
Our current understanding having spoken to systemd developers is that we should be able to find a path that brings us much closer to upstream, if not entirely.
The only way the systemd developers will allow musl support upstream is if musl supports the glibc-isms that systemd uses.
They have been extremely clear that they will not carry patches for other libcs.
Makes sense