Can you give some examples of implementation that are hard to understand and use?
- mediocre config files - they decided that config files are declarative, which could be good, but it pushed simple (and visible) logic in /etc/init.d/* files into more complicated logic encoded in binaries. The simplistic config files seem to be more simple for systemd to implement, but harder to understand, maintain and use.
- binary log files. everything has to go through systemd to access them.
- responsibility creep - systemd started taking over more functions, and it doesn't seem to stop. Will systemd turn into busybox, with simpler poorer implementations of every single thing in the system? dnsmasq is similar. The thing is, busybox and dnsmasq are cut down implementations for smaller system that are alternatives to their full-featured brethren. systemd takes over the core of a system.
That's fine, and I'm OK with a world where we disagree about how usable systemd is, but I interact with the systemd config files and symlinks and logs pretty frequently, and have found them to be orders of magnitude more usable and understandable than /etc/rc#.d and /etc/init.d and "I wonder if this system puts sshd logs in /var/log/auth.log or /var/log/secure or somewhere else". Notably, I interact with a mix of Arch and Ubuntu and RHEL systems over the course of a week, and systemd means I can engage with all of the above using the same methodology.
In terms of creep... busybox is noteworthy for packing all of a system into a single binary. systemd doesn't do this. Their resolver and time daemon and networking and all the other bits and bobs are their own separate binary, and you aren't required to use them.
systemd is functional. It does the things you say. It unifies things, including config files and log files. It works for you.
I just don't think the way it did this is that elegant.
My busybox analogy should be more viewed as: what if arch/ubuntu/rh decided that busybox would replace all coreutils, etc on their systems today? All the same commands. Still usable. But something is missing...
I have the feeling that you are mourning the old, bad days?
It is actually way better understandable and manageable: /usr/lib/systemd is for distro-supplied and packaged units. /etc/systemd is for your local overrides, /run/systemd is for units created at runtime by various generators. (Then there are $XDG_(RUNTIME|CONFIG)_DIR/systemd for user units, for consistency).
So no more diffing config files at update time: the distro provided ones are always updated, your overrides are always preserved, cleanly.
> - binary log files. everything has to go through systemd to access them.
So send them to external syslog, that saves them as you wish. There is a point in indexing them (and being able to have a newline in the log message).
- responsibility creep
This is a matter of perspective. If you take it as "init system at boot", then yes, it might looks so. If you take as an service manager that runs/stops/reloads/monitors units, based on different events (startup, dbus activation, socket activation, timer, etc): what's the difference in the bulk of action, except for the event that caused that action? The service handling, unit definition, etc. are all the same, just the initiation event is different. In this light, it doesn't take over functions; it consolidates and unifies them.
Unlike dnsmasq or busybox, it is not in cut-down fashion, but exactly opposite. It replaces rag-tag of tools with one with more complete and tested implementation.
check out for systemd networking, the resolver, timesyncd, etc...
for example timesyncd IS cut-down and brute-force:
"does not bother with the full NTP complexity, focusing only on querying time from one remote server and synchronizing the local clock to it"
ntp is complex, but it does many things in safe and effective ways.
etc.
systemd-resolved, on other hand, is good. There's no other resolver stub, that has the functionality that resolved has. It is the only one for Linux, that can do per-link, split DNS properly.
Addendum: https://news.ycombinator.com/item?id=15228940 This approach clearly deviates from widely expected behavior, whether it is in conformance with applicable RFCs or not.
See, many people have misconceptions how it works; see the comments in thread you linked to. Add to it the fact that resolv.conf syntax cannot express what is expected of modern resolvers (not even with apple-like extension - `/etc/resolver/*`; it cannot react to links going up and down), it is basically just compatibility shim for broken applications (they are not supposed to use it directly); that's why systemd-resolved configures it to point to itself and handles the specifics for them.
That understanding has been wrong; it is an implementation detail of glibc and that's it. It could query them in random order, or round-robin them for each request; or time them how fast they respond and prefer the fastest, up until it starts timing out and then shake the order again. Some resolver stubs ask all of them and use the answer that arrives first. That's also valid. Since they all are returning the same answer, it doesn't matter. It also makes any argument for ordering superfluous.
So that "common understanding for many" is erroneous and those many should update it. The thread you linked to has exactly a discussion on this point, so in the years that passed since then, those affected many had ample time to fix up their assumptions.
Again, there can be other valid reasons for maintaining an order. Could be about latency, privacy etc.
I admit to not having read the full thread, and I apologize for any needless rehashing of arguments. It was just the first result confirming systemd-resolved breaking setups by its handling of what many (in that thread) consider expected behavior.
What I think you're primarily suggesting is that there are no formal standards prescribing glibc's behavior in terms of DNS resolution order. That is certainly true, and I don't disagree calling it an implementation detail.
1. symlinks: I think the systemd structure is very intuitive and the order of overwriting makes sense. It's new, but it makes sense, and we only have 3 different locations, each with a specific order and precedence.
2. Config files: I think keeping the declarative syntax is the secret sauce of systemd. The whole point is you can orchestrate processes without needing to write code. They're much easier to understand if you assume a knowledge point of 0, and they're much more maintainable (no logic), and they're even more portable. Bonus is they're much faster to execute by the init system. Yes, most of the logic is in systemd now, but you tell it what to do. Reading the documentation goes a long way to making unit files less cryptic.
3. Binary logs: yeah I agree just not a great idea. Always sounds nice on paper because you can do neat things with permissions and can have faster performance, but in reality just not necessary. They got too fancy.
4. Responsibility creep: from where I'm standing, not a problem. systemd isn't a piece of software, it's just a name given to a collection of software, like GNU. systemd-init doesn't have feature creep. Sure, now we have systemd-file-my-taxes, but that's optional and it's a whole different program that just uses the systemd name.