About Devuan

qyron · 2017-09-20 16:15:22

Forewarning
I do not have the intention by asking what I am about to ask to start a flame war nor in any way to disrespect any or all of those involved in any form with the release, maintenance and development of Devuan.

What are the future prospects for Devuan?

I came to the Linux world in search for something that offered me what I had lost in terms of OS reliability and I discovered it in Debian. Lately, with the emergence and establishment of the dreaded systemd, I've read many warning against it. I will not lie and say I fully understand the ramifications and deep changes it brings to the OS: I do not. Although very fond of computers and technology, truly dive into the reality of OS architecture would take time I do not have to spend. I understand systemd introduces deep changes that for many are a breakaway from the original guidelines for Linux, besides being in itself a vague and ill formed proposal for advancement, with too many vague and opaque areas to be considered reliable. Another offshoot of it is allegedly making the system heavier than it needs to be and breaking away with older hardware.

To my understanding, Devuan at this point aims at only purging the original Debian structure from its dependency of systemd, with further development and distancing to come in the future. I respect this. But I also understand that Devuan, because of this, is behind the Debian release cycle, because of the need to work on the present version in order to purge it. As I understand it, this is why Devuan 1.0 is tied to Debian 8, while Debian itself has already advanced to its ninth generation.

This brings me to my next thought: I'm all about stability, but being behind the release cycle of Debian, implies that the userbase of Devuan will be getting dated software (I have read that Devuan already ships the latest version of XFCE, so my last sentence may very well be garbage, but allow me to finish my train of thought). Older software offers better stability but too old becomes counter intuitive, if the intention is to attract new users, as most users tend to want modern software. Does Devuan intend to break farther away and catch up on Debian like other distros have done, by working from the stable/testing branches directly, just siphoning the software, or will this delayed release cycle persist? I've already read that even the sysvinit is on its way out, so I think I can see in this a deeper breakaway from the the Debian tree and further changes forming to the future.

But right now, besides the ideological stance (which is precious by itself) what does the average user gains by migrating from Debian to Devuan?

Thank you your attention and I sincerely hope someone will be able to shed some enlightenment on this.

fsmithred · 2017-09-20 19:44:34

Faster boot times. (no more 5-minute waits for a start job)
Predicatable interface names. (I know it's going to be eth0 unless I changed my hardware.)
An init system that's been well-tested. (I might switch to systemd once they're done developing it and it's had time to mature.)
An init system that knows it's an init system and is OK with that.

Probably more, but it's time to let the dogs out.

qyron · 2017-09-20 20:18:39

Thanks.

I'm really considering trying Devuan out. I just migrated from 8 to 9 on Debian. I usually just upgrade when a neew version comes out; I'm not about distro hoping. I choose one and keep loyal to it.

What is holding me back at this point is the age of the software.

fsmithred · 2017-09-20 20:24:23

While it took two years to get a final release of devuan jessie, it was usable long before that. A lot of the delay was just some loose ends to tie up. (Couple of problems came up in the installer.)

We don't expect the next release to take that long. If you test it out in a VM or on a spare computer or even one of the live isos, you'll feel like you're using the debian you remember.

greenjeans · 2017-09-20 20:56:41

qyron wrote:

What is holding me back at this point is the age of the software.

Long term support from Devuan for jessie, i've heard until 2020 or so.
LTS kernel 3.16 too, projected end-of-life is April 2020

Software age overall doesn't mean much except for certain newer hardware that's been extensively microsofted to try and keep us out, lol, I understand totally if you need new drivers and firmware etc.

I use gpicview for a picture viewer on my systems, hasn't been actively developed since 2007 when the dev abandoned it and recommended everybody use something else since he wouldn't be adding to it or supporting it anymore.

A decade later and still nobody has made anything any better, it still works perfectly, lightning-quick and light on resources.

There is such a thing as homeostasis in a piece of (simple) software, when it reaches a point where it does everything you could ask of it, has no bugs, and just works indefinitely with no more maintenance required.

And that's about the time the developer gets bored with it and looks for some new stuff that's buggy as hell and overly complex, because that's more fun.

{...sigh....}

FOSSuser · 2017-09-22 09:18:58

Well, Debian software is behind mainstream distros anyway, Devuan will be on a par with Debian in the next release.
It just took time to purge the system, initially, of the entangled systemd.

Systemd is proposed to work like MS Registry, controlling most things from one encrypted file - totally against unix & linux principles.

foobar · 2017-09-23 06:57:53

@FOSSuser: Instead of packaging stuff Devuan packagers prefer to fight over the next iteration of their build infrastructure. Nobody has any idea where things should be headed, the current infrastructure is unmaintained (e.g. the Devuan gitlag is unsupported upstream by now!) and new people are being demotivated at every turn. Just ask golinux about how her commendable effort to increase communication in the project was shot down. No developers are around that can actually make sense out of C code.

I am not at all optimistic about devuan anymore.

fungus · 2017-09-23 09:01:36

I think that Devuan gave it one hell of a shot. It took the impossible, jessie w/o systemd, and made it real. The route as I understand it is either to fork its own distribution with what it already has been able to cleanse or progressively fall further and further behind Debian. When systemd progressively expands its domination, in a distribution supported by thousands of developers and package managers, it is not ability or good will, it is man-hours. When you have thousands intentionally destroying, it will not matter how well and how fast your small team builds. The outcome will be total destruction.

If devuan developers were counting for a surge of funding of their revolutionary project, I think they are naive idealists. And I am sorry for their expected disappointment. Money goes with money, interests go with money and interests, and that is h/w manufacturers, consulting, and "control of information". Either they will vanish or not do what they have been doing.

What do we have? A good base, a springboard, to do things on our own. Are we willing and able? Who prevents any of us to switch a functional ascii or ceres to debian only repositories, block out "apt-get dist-upgrade" from ever running again, and simply do debian "apt upgrade" of what already exists and works, while keeping our own cache/repository? If something comes up with an non-existing dependency get the source, hack-it patch-it, and go. Hopefully share your success and maybe collectively we can go a step further.

Out of curiosity I have a buster and sid distributions and three days of being unattended had a storm of updates. It is not just updated packages it is "new" packages and "purged" packages that make those distributions functional.
Devuan and Refracta have provided many of us the ability to fork our own projects. The torch has been passed and we should be thankful of those three-ten years of hard work that was handed to us. It is not over till we say it is over. Manjaro, just like Ubuntu is sold pre-installed now in expensive hardware. The easy way is sweet.

If you have a consumerist kind of attitude demanding better service and support chances are you were here for all the wrong reasons.

The struggle must go on at all costs. Otherwise we should be cut off from the grid and head for the hills on foot, if you know what I mean.

Last edited by fungus (2017-09-26 07:30:15)

catprints · 2017-09-23 13:40:35

I am optimistic. Stable was released in late May. Ascii soon to follow. Instant gratification is foolish to expect.

I suspect fud is at work here.

Last edited by catprints (2017-09-23 13:49:17)

greenjeans · 2017-09-24 21:05:16

FOSSuser wrote:

Systemd is proposed to work like MS Registry, controlling most things from one encrypted file - totally against unix & linux principles.

Not for nothing, but I never see anyone mention dconf, but it's essentially the systemd of config and was there before systemd even got here, to me it was really one of the first big shots fired against the unix/linux principles you mention, and still is.

A binary config file in userspace ,that you can't read, that controls every bit of configuration in your system, and will in some circumstances replace ALL your configuration with some basic crap it dredges up that it thinks is best.

It's like hitler and sauron got together and decided to forge one config to rule them all, but they were drunk so what came out was just a village idiot on meth driving a bulldozer.

IMO dconf needs to go bye-bye almost as badly as systemd does.

fungus · 2017-09-25 07:39:39

where is the LIKE button when you really need one!

@greenjeans, I have promised mother @golinux that I will behave if I am to stay here so ....... no comment

golinux · 2017-09-26 01:35:45

golinux sends fungus a virtual cupcake with a sparkler on it!

fungus · 2017-09-26 07:24:49

Can you export virtual cup cakes out of the Netherlands without virtual dogs sniffing them?
I thought I'd lure you in taking the stand on this discussion, but you resist.

I did a little research on dconf, and there exist such a thing as dconf editor, some gnomy gui clicky thing.
The "funny" thing is, it does not edit all that dconf controls and there are warnings that editing can break
your system. It is like playing with the wheel of an airplane with autopilot on. Only the button that turns
off the autopilot can affect the plane.

I am infatuated with UUID (as a necessity) and the ability of iPv6. Intercepting a few packets can send
special forces to your door and live evidence of who said what. It is freedom of speech after all, not freedom
to write in public.

Last edited by fungus (2017-09-26 07:25:36)

Racoton · 2017-10-16 14:08:07

My idiomatic simplicity just let me say: this thing is growing now. And I'm glad. Long life for Devuan.

The officially official Devuan Forum!

#1 2017-09-20 16:15:22

About Devuan

#2 2017-09-20 19:44:34

Re: About Devuan

#3 2017-09-20 20:18:39

Re: About Devuan

#4 2017-09-20 20:24:23

Re: About Devuan

#5 2017-09-20 20:56:41

Re: About Devuan

#6 2017-09-22 09:18:58

Re: About Devuan

#7 2017-09-23 06:57:53

Re: About Devuan

#8 2017-09-23 09:01:36

Re: About Devuan

#9 2017-09-23 13:40:35

Re: About Devuan

#10 2017-09-24 21:05:16

Re: About Devuan

#11 2017-09-25 07:39:39

Re: About Devuan

#12 2017-09-26 01:35:45

Re: About Devuan

#13 2017-09-26 07:24:49

Re: About Devuan

#14 2017-10-16 14:08:07

Re: About Devuan

Board footer