can systemd do ANYTHING other systems can not do? Can you transition to wayland?

NO!

NO, what? To which question, both?

Then what is the problem with systemd or replacing it with another system?

The quick answer is NO, but in reality one must be able to match the resources of IBM/x-RH to chase behind this dog when this dog does its best to try to shake all tailgaters away from it.  “It” wants to run throughout the system, from bios/efi/boot-loading to every click you make on the desktop with your mouse or finger touch to the touch-screen.  PRECISELY what google has done on android, for which we have no say it is a private non-open non-free system.  Monkey see, monkey do, monkey governments enforce its use, monkeys conform.  Fuck you google and fuck the state that mandates its use (I wonder if North Korean sites have google apis in them, RT do!).

HOW does this work?

Due to the race of popularity (which translates to donations contributions and attracts younger hopeful contributors) of desktops and desktop specific tools and gadgets, systemd provides more and more toys (functions) for them to use, instead of writing more code of their own, or using more generic functionality that all systems can use.  The more those tools are used (just like Google’s model of providing free cgi tools for websites and web-uis – without google most sites fail today, even this crappy wordpress.com site to operate anymore it MUST use gravatar, gstatic, google-analytic – the more control systemd takes over the operation of the desktop.  Are those functions irreplaceable?  No, it just takes time (resources) to replicate your own system’s (open-rc, runit, s6, …), and not only create the functionality/mechanism but also provide a library name equal to systemd’s that does the same thing.

With this in mind we have IBM fan boys, and facebook fanboys, say “but the code is good, it is open, it is free” and I will use it.  Within weeks zstd was to replace xz everywhere, magically, no incentive, even though inferior.  Allow some time and you can not possibly have a distribution without zstd.  Even the linux kernel defaults in compressing/decompressing modules with zstd, codecs for browsers and media players look for zstd.  The first argument in defense of zstd’s performance was it can use more processors.  Now, after a few years of use it has adopted the -T# tag to specify how many threads/cores to use JUST LIKE XZ did!  But did any distro pre-configure xz with -T0 ?  No, they were too busy defending their “choice” of facebook’s code (facebook purchased and published the code as free!).

Can you stop your users for demanding the latest version of browser, of a messenger, of this that and the other, because you don’t have the time and man-resources to reproduce the new functionality, or even the renamed functionality so this new crap will work?  No, you either shut down or you adopt entire pieces of systemd (as cgroups, elogind, esysusers, etmpfiles, e-udevd or xdevd they call it, systemd-boot, and now there is more, basu!).  If you remain stubborn, users will install the same from flatpak which may incorporate a ton of crap just to run the “app”.  And the system will turn into the carnival of xyz.exe that runs it all!  We already have isos out of full desktops that equal or exceed microsoft win11 installation images.  Wonder why, since you have nothing better to do.

So it all boils down to resources, money and knowledgeable soldiers to execute the orders from some captain of the project.  If the naval commander says OK, incorporate more systemd/IBM code and functionality and please the “customer”.

BASU is an other piece of systemd, necessary as building and running dependency for some wayland related gadgetry.  You can get a wayland compositor running, such as labwc (the best I have encountered), that simulates openbox just with seatd.  Sway doesn’t play well lately without systemd/elogind.  It is getting hardcoded, and this may not be sway itself but all the sway tools any wm would need, in lack of alternatives.  So to get some innocent window manager in wayland working may be easy, but many utilities commonly found on X such as xrandr/arandr don’t work.  The only ones available are those made for sway, that may or may not work with other window managers.  You can run feh as a background setter in X11, and since libx11 xorg-wayland promises to run X applications in wayland you think it works.  NOPE!! Not yet at least, but swaybg works – it keeps a process running just for the background setter.  Maybe the only thing that can set a background in a wayland compositor is swaybg, most others revert to swaybg to function.  So here comes basu, hardcoded in some wayland utilities, and you say, what the heck, in absence of an alternative I will use basu!

So what do we here in sysdfree have as criteria?  On the strict list we exclude systemd-free distros like artix and void because of they use elogind, but  if they use basu we are OK?  I don’t seem as flexible as accepting something like this, unless there is a very convincing argument I haven’t considered.

Kiss, a long time ago, said we only support wayland and it doesn’t need systemd.  Gradually it seems it does, not directly, because your goal is not to be able to run a minimal window manager but to do work on it, and  therefore you are lured in like a good little fish back where they wanted you.

I am wondering now whether you can even compile basu in musl, does alpine or void have basu in their musl repository?  If it does it is probably portable to BSD, BSD desktop addicts will use it … and upstream the green light to directly code systemd tools, functions, utilities in just about anything they do.  xbps-query -Rs basu : SD-Bus library extracted from systemd

Does s6 have a single distro to demonstrate it works?  NO!  It has a few that adopt it in slow to faster rates after the changes have been consumed and digested.  Version after version, if you dare write code over skalibs, execline, s6, s6-rc, a new edition will break your work.  Unreliable and monolythic?  It is either skalibs/s6/s6-rc/execline alone from master or the highway.

The future looks very hazy and full of uncertainty in all fronts of alternatives because of ONE reason.

The reason is what we call open and free.  And since free can also be provided by a corporation with nearly unlimited resources, ALL FOSS can become a dog chasing this evil corp.  The value and principle to distrust corporate code and seek simple alternatives to anything they furnish you is not there yet. 

But it is just a computer system, not political philosophy! 

If you dare consider that a for-profit corporation will spend 4m of a light-bulb to stay on so you can finish your open-free code without seeing a benefit (profit) you are so disillusioned no drug can be worse or bring you back to reality.  They will not even waste soap to clean the employees bathroom if they can not see profit indirectly linked to it.  It is how this cookie crumbles (I wish it did).

You are also disillusioned if you think that corporate for-profit interests don’t hide behind other projects who are published by “non-profit” manned by employees who make 6digit salaries (mozilla, qt, etc.), and that Red-Hat wasn’t IBM’s trojan horse all along.  You expected the official announcement to accept it.

So what do we do here?  Republish a strict list of non-systemd distros under strict standards, till when, till there are only 2 or 3 left, and those 2-3 only have about 600 packages and nothing else will run on them? 

If we add basu to the criteria some will cry ….  they were confronted with blackmail and extortion by distrowatch to admit them or drop them, now to be dropped from sysdfree as well … ?  Distrowatch now requires a distro to have its own forum to discuss matters.  So void has a reddit board.  Reddit is turning to such a crappy web-ui that you can hardly find a way to publish a code box in this mess.  Reddit started from people discussing and sharing ideas about code, but now it is just a general interest social medium.  Madness …

This was meant as a community project, so come on community, speak up, what do we do?

It is not a dead end yet, we can stay on x11, not use elogind or other logind, no dbus, no seatd, and things still work, for how long I don’t know.  Nobody is forcing us to wayland, nobody is forcing you to update your favorite browser either if the next upgrade breaks.  Webpages will keep breaking, a/v media will be inaccessible, and all we have is glibc and a terminal 🙂  rxvt

Not in the greatest mood to write an article with optimism, but it is what it is.

Any optimists here?  Speak up, we need your help!

 

 

9 thoughts on “can systemd do ANYTHING other systems can not do? Can you transition to wayland?

  1. You should distinguish between standalone and forked projects. You could inquire with the Basu maintainers to see if they adhere to systemd policies. I did, and received a clear response: ‘We do not strictly follow systemd policies. New code from systemd is not automatically integrated, and we have the flexibility to either develop our own code or make improvements to the internal code for API enhancement.’ Basu is a fork, not a standalone program like elogind.

    Basu provides API libraries for C code to communicate via dbus. This libraries was initially developed by systemd developers (this is a fact), but it does not mean it is unusable. The challenge here lies in the time required to develop a robust dbus API library. Creating such code can take hours, weeks, or even years of development time. Is it ethical in the FOSS community to use code that was not written from scratch?

    It’s regrettable to feel ‘forced’ to use ‘evil’ code, but this does not imply any dependence on Systemd philosophy, politics, or policies. However, it force you to be dependent of Dbus API interface changes.

    There are other examples of code to consider. For instance, the ‘libnih’ libraries offer better integration and are less complex than systemd, in my opinion. But it’s a old one and surely need improvements. Additionally, consolekit provides dbus API libraries.

    You should ask why dbus?

    Maybe you can’t remember the “HAL” time ;).

    Communicating between different processes with a common interface is a good concept. The way is was made can obviously be discuss but the concept by itself is, should i say, needed. Allowing programs to communicate with the system is a great features, and system reacting on the request is also a great feature.

    Like

  2. Hey, throgh from Hyperbola here: If there is interest for reading … Hyperbola has thrown out many Linux-only frameworks. There is also no wayland included and we plan further for removal of eudev. I leave now the link here for a collection of ideas for 0.4.6 and beyond: https://forums.hyperbola.info/viewtopic.php?id=1028

    We really need helping hands for exactly that step of eudev-removal and patching several base packages. So if there is interest? 🙂

    Thanks for the article.

    Like

  3. Please add a nickname in the box below the comment or at least use a nickname in the comment text so we can separate all the anonymous comments, it is confusing when two anonymous comments conflict each other.

    The point about this being a community is that through discussion we can convince each other what is right and wrong, good and better or worse.

    OK! I stand corrected!

    Basu is a fork NOT an extracted piece like elogind etmpfiles esysusers ….

    eudev and consolekit had once been parts of the code and forked. 

    Read the proposed patch needs below

    https://github.com/ConsoleKit2/ConsoleKit2/issues/150

    https://github.com/ConsoleKit2/ConsoleKit2/issues/151

    This is both stemming from sway and its need for basu, if not the entire caRnnibal.

    As time moves on and more projects (Plasma) move to wayland and AWAY from x11, more and more projects will move. More and more window managers new to wayland will rely on already developed tools, and the concentration momentum has sway in the center. The most convoluted, large, hyped wm of them all. 10-15 times bigger than some more stable tiling managers, even floating or both.

    The great work done by labwc to simulate openbox and make transition seamless may be great, but basic toolsets are not provided, and are sway byproducts. What do you use for xrandr? Sway tool in need of basu. What do you use for fancy customizable menu or bar or dock? 

    The more people rely and adopt sway tools the more dependence to this caRnnibal will become prevalent. The aim clearly, by IBM, is to kill ALL alternatives and concentrate dependence on its OWN!

    In that respect, basu, consolekit2, eudev, are all doomed underpowered projects asked to run at IBM speeds. One doesn’t have to be a palm or crystal ball reader to see there will be many dead dogs in that trail of the caRnnibal!

    We have mdevd and a real need to bridge the gap between the skarnet absolutism and the libudev-zero realism. Basu is just a tiny little initial trojan horse logind once was. It is meant as lure, then it becomes the bridge of a ship, then there will be many drowned projects left behind in open sea.

    Like

  4. Eric here (the very first comment was also me). Actually, eudev is not a fork but a standalone program. So, they follow systemd policies decisions.
    For now, I haven’t seen any real alternative to udev. Some projects are attempted but they cannot be used as replacements.

    Knowing the hardware of your system isn’t that difficult. This was achieved cleanly with mdev.
    The real challenge is to notify and react to events when something changes on your hardware. This implies that you need to communicate with “udev” and programs in both directions. To do so, you need a standard syntax that any developer can use to communicate with “udev,” and “udev” needs to respond in a standard, understandable way. This is actually the missing part of the udev alternative. I think taking the base of mdev and creating a wrapper around it to translate configuration files from mdev to udev (and vice versa) can be a good approach.

    Some interesting projects was begans about logind replacement : https://github.com/chimera-linux/turnstile. Thanks to q66 to initiate this.

    The aim clearly, by IBM, is to kill ALL alternatives and concentrate dependence on its OWN!

    This can also be observed in a professional ecosystem. It’s not for nothing that RHEL 10 will only support Wayland.

    Like

    • I don’t get it: I’m running mdevd here on a full desktop xfce, and have more than everything I need. mdevd notifies, and reacts to changes in my hardware in real time, and it does what *I* want it to do (or rather, what I *tell* it to do), *NOT* what my installed apps think it should do.

      I neither need nor want installed software to interact with mdevd without my intervention. So, what am I “missing” due to this?

      Like

      • Remember systemd is IBM, some say just recently, I believe it has been since day 1 but in case it backfired in the development why associate the name with it. HW corporations, and propriatary software MUST collaborate and defeat common enemies. You never know what HW will come out to make your mdevd more disfunctional so you will rely on “their” subsystem. This nvidia crap with high performance marketing sheets, is nearly 50% of all user problems addressed in linux. Why because of their infinite blobs they must convert to “plugs” for linux to work around. Look at all the non-PS printers and the blobs trying to make them work with linux/cups. I have seen printers that say “linux” compatible on the box, and what they mean is debian/ubuntu *.deb firmware unable to even use in arch or manjaro. USELESS!! You scan them over with android pos (crap) and beep it prints.

        Not coincidence!

        Like

  5. Pingback: the rise of basu through the maze and into the core of build-chroot – how? | systemd-free linux community

If your comment is considered off-topic a new topic will be created with your comment to continue a different discussion. This community is based on open and free communication, meaning we must all respect all in minimizing the exercise of freedom to disrupt such communication. Feel free to post what you think but keep in mind the subject matter discussed. It is just as easy to start a new topic as it is to dilute the content of an existing discussion.

This site uses Akismet to reduce spam. Learn how your comment data is processed.