[
Can you do that with sysvinit?

Re: init diversity in Debian

Postby HuangLao » 2019-12-29 16:24
Well, I've been following this thread and a few others as a lurker and now feel "compelled" to write.

I find a lack of sincerity or perhaps a bit of hubris in the apparent alarm and surprise of the recent GR vote regarding systemd. We know from the prior vote that Debian decided to default to systemd and to follow what other distros had already done. This vote simply reinforces the first vote, however, it does not close the door on other init's, it merely says it will not be the primary focus of Debian as a project nor will it enforce that policy on the dev's. If a dev wants to support other init's they may, if someone wants to step up and do the work they can, they may accept patches to support alternatives and they are encouraged to discuss ideas rather then ignore or brush off etc... As fsmithred stated:

    I don't forsee any immediate changes. Yesterday's vote leaves us in a slightly better position than before. There was no commitment at all to work with us. Now there is. It wasn't my favorite proposal, but it's something we can work with. It could have been worse.

https://dev1galaxy.org/viewtopic.php?pid=19182#p19182

and anticapitalista:

    Let's hope that those packagers that do include support for alternate init systems besides systemd in buster will continue to do so.
    We at antiX have found buster to be much better in terms of init-diversity than stretch was.

https://dev1galaxy.org/viewtopic.php?pid=19185#p19185

Keep in mind the historic nature of Debian, it is a very slow moving ship, no-one ever said it does not follow other trends it simply gets there at a much slower pace, and typically does leave room for slight alterations of course. Is that not what the recent vote concluded? Continue to move towards systemd as primary but remain open to alternatives.

Course forward if you base your distro on Debian or any systemd distro:
1) rebase on a non systemd distro, start the work now for your next release

2) continue basing on Debian and fork or rebuild any programs that require systemd, the source is out there.

3) block any programs that need to be rebuilt or cannot be rebuilt because they are too "tainted". Slackware stopped providing Gnome many years ago because it became an abomination to maintain.

4) become a Debian dev. so you not only can contribute to the upstream of your distro but also gain a vote/speaking voice in the direction of that ship, this is what many Ubuntu dev's did a few years ago and what Debian dev's did for Ubuntu prior to that.

5) unify the several small distros with a common goal around one common distro/mission....example, antiX, MX, Devuan could combine together either with some work or all

6) propose a systemd free repo for Debian similar to the nonfree and contrib repos, that is maintained by antiX, MX, Devuan etc.... Of course to make that proposal you would need to be a Debian dev...so it goes back to that.

Finally, stop overreacting, the sky is not falling, Linux is not dying, challenges like this come up every few years and it will in a few years again. When you have been around the game long enough, you stop reacting to the infighting. PV/Slackware is just now exploring adding Pam/Kerberos many years after it became default in every other distro. You do not need to rebase off Slackware to learn from it. I do not think (cannot speak for him) that PV would be interested in a merger, even though PV and Ian did ponder a merger early in Debian's history. It was actually Ian wanting to merge with Slackware. With that said SalixOS may have interest but you would want to speak with George (gapan) about that. Slapt-get has apt like functions and the Salix team tends to be more open to Debian and other distros (Arch etc...).

Slackware probably still avoids systemd.

I do not care anymore. IMO, It's not too late to think of any other non-Linux system. I'm already practicing my hands with BSD. I'm looking at Haiku aswell. Instead of learning another Linux OS I find it more reasonable to use a non Linux OS.

Curiously, I installed Debian 10 a couple of days ago and didn't hold it for more than a few hours. Still no to"C:\SystemD" for me.

Now I am wondering if I ought to open a completely different thread. Before resorting to trying a completely different distro to escape the random freezing problem, I decided to experiment with enabling swap. Since modern computers always have plenty of RAM, at least for ordinary desktop use, I have tended to skip the swap partition and just run any Linux without swap at all. But when I edited a swap partition into my Devuan test system and enabled it, things seem to work much better. Haven't had the time to test extensively yet, but it seems to have run for a whole afternoon without any freezing. This seems bizarre, since I've never come close to filling the RAM on this machine, so I don't see how swap would make any difference. But as I have searched around other forums, the discussions (which can get very technical about kernel issues) seem to say that swap is very deeply embedded in how the kernel works and you ought to have at least a token swap partition on any system.