- This topic has 7 replies, 6 voices, and was last updated Oct 26-1:03 pm by techore.
- Posts
mcgiwer
Hello. I would like to propise adding more variants of the init app’s to choose from.
As is widelly known, each init has it’s own pro’s and con’s.
Adding more choice options would allow people to suit better their needs.
I had read that OpenRC is also a good alternative of it.
- This topic was modified 1 month, 1 week ago by mcgiwer.
Karo
I’m not part of the developer or maintainer team, but I think it’s difficult. Someone would have to do the extra work.
Original german text:
Ich gehöre nicht dem Entwickler oder Maintainer Team an, aber ich halte das für schwierig. Irgendwer müsste die zusätzliche Arbeit leisten.Since english is not my native language and my knowledge is too poor to represent complex things I use a translator.
These are known not to be perfect. If you translate one of my posts from german to english into other languages, the result will be worse and worse. Therefore, I post in both english and german, so that non english speaking users can have the german text translated directly into their language.
Brian MasinickSome time ago when we were having discussions on the forum about what *might* be some ideas for a future release, anticapitalista himself mentioned that he was considering at least two alternative init alternatives. As I recall they were s6 and dinit.
I caution you – this was only a discussion; there were no dates, there was not a definite promise this would happen, although given the current situation with sysVinit it may become important to eventually “move on”. Unless s6 is implemented (and that’s probably the likely choice IF there is a change), sysVinit won’t just “disappear”; we were assured of that quite a few months ago. Certain services just don’t run the same way in runit, at least as it was during the testing cycle prior to the release of antiX. If and when there are changes, they’ll be coming from anticapitalista, not me, and not before there’s a LOT of work and consideration.
Here are some resources for those who may be interested in looking at some of the alternatives.
According to https://www.howtogeek.com/713847/the-best-linux-distributions-without-systemd/
“s6-Linux-init: s6 is a replacement for SysV init that tries to address the serial nature of SysV init and remain true to the Unix philosophy.”https://www.slant.co/topics/4663/~linux-init-systems mentions a number of init alternatives, including dinit.
https://davmac.org/projects/dinit/
“Dinit is a service manager / supervision system, which can (on Linux) also function as a system manager and init. It is:
Portable — written in portable C++ code; compiles and runs on a variety of Unix-likes (Linux, various BSDs).
Compact — with a reasonable feature set, but a sufficiently limited scope; designed to work with other system components, rather than replace them.
Robust — written with a focus on being secure and correct.
Usable — with simple configuration syntax and control commands.
Well documented — check the extensive manual pages!
Free and Open Source — distributed under the Apache License version 2.0”More at https://github.com/davmac314/dinit/blob/master/doc/getting_started.md
--
Brian Masinick
sybokHi, not an developer but adding an init alternative means preparing the init scripts for it.
Debian uses systemd by default; systemd will drop (or it already did?) sysvinit compatibility.
This means that maintaining sysvinit may become harder if the corresponding init scripts are no longer maintained and phase out of Debian.
This could be an opportunity to consider switching from sysvinit.Currently, official antiX ISO releases offer: runit and SysV init ISOs.
There was a rather long discussion before releasing the antiX-23 runit ISOs about the runit services (very good points provided by @abc-nix who was explicitly mentioned in the release thread, please see https://www.antixforum.com/forums/topic/antix-23-released/#post-115034).
This means that getting the runit work (more) correctly was quite a lot of effort.Even when re-using and adapting init scripts from other Linux distro (e.g. MX Linux if they decide to go this way too), this would be a lot of work as already mentioned by @Karo.
Not shooting the proposal down, simply pointing out related context and effects of the proposal.Xunzi_23
How about A Poster learning to make the alternative inits work with antiX then making a post along the lines of.
I am xyz and have made a fully working antiX version with OpenRc, if others are interested in further testing bug
hunting download of an ISO is available from my page on: Git or other choice… Checksums are also available to
ensure download is correct.AntiX is very much the work of the few for the benifit of many…
I also worry about debian backstabbing on alternative Init as well as further continual action by systemD dev and packagers
trying in every way possible to damage alternatives to the holy pottering and IBM red hat cause. Debian moves ever closer to
Red hat linux with apt rather than staying independent.Wie wäre es, wenn ein Poster lernt, einen Alternativen Init einzurichten und zu nutzen. Zu Testen und Sicherstellen dass es mit antiX funktioniert.
Danach ein Forum Beitrag erstellen.
Ich bin xyz und habe eine voll funktionsfähige AntiX-Version mit OpenRc (S6/Dinit) erstellt, falls andere daran interessiert sind ein ISO ist von meiner Seite auf verfügbar: (Git oder andere Auswahl.).. Prüfsummen stehen auch zur Verfügung. Bitte helft mir mit Fehlersuche und Bugfixes.
AntiX ist das Werk einiger weniger zum Wohle vieler… Ich mache mir auch Sorgen über Debians Verhalten, bei alternativen Init-Angriffen sowie über weitere Aktionen von systemD-Entwicklern und gläubigen die versuchen , auf jede erdenkliche Weise Alternativen zum heiligen pottermann und der IBM Red Hat SysD Kult zu beschädigen. Debian rückt immer näher an die Red Hat Linux mit apt Distribution, anstatt unabhängig zu bleiben.So nebenbei ist ein Online Übersetzung von Englich auf Deutsch völlig sinnentstellende Müll gewesen. https://duckduckgo.com/?q=translate&ia=web unbrauchbar….
when I get a possibility then I could try to create ISO with openRC to test
Brian Masinick@xunzi_23 wrote: “How about A Poster learning to make the alternative inits work with antiX …”
I agree completely. When runit was initially introduced, @Xecure was alongside anticapitalista, helping and setting up as much as possible. I believe that Void was used to a certain extent as a model, but model or not, it’s a huge undertaking to put any new or different init in place.
In antiX 23 @abc-nix helped a great deal to further improve the procedures. I understand that not very many people have either the time or the skill to do this type of work, so whatever does happen, IF it happens, is going to need plenty of cooperation, and when it’s ready to be tested, even if we can’t develop, we need a LOT of people to test; otherwise we’re going to find out stuff when a new technology is deployed, and that’s not fair to the development team.
At this point all of this is pure speculation, though at some point, assuming we want more than one init alternative, something will be needed, because if I understand correctly, support for sysVinit is not planned to continue much longer (if it’s there now). I’ll see what I can find out; if any of you know or recall any other details, please share them; thanks!
--
Brian Masinick
techoreI don’t have anything against OpenRC but our implementation of runit is incomplete. Before tackling a new init, let’s work on expanding runit. My two cents for what it is worth. 😀
- You must be logged in to reply to this topic.