Skip to main content
Topic: Artix "philosophy" in terms of the system (Read 2034 times) previous topic - next topic
0 Members and 2 Guests are viewing this topic.

Re: Artix "philosophy" in terms of the system

Reply #15
TL;DR: Don't expect any kind of deep philosophy from Artix any time soon.

I switched to Arch from Gentoo in 2004 solely because I was tired of compiling (and fried a laptop in the process); the amount of freedom Gentoo gives is indeed second to none. However, the way of Arch back then was fairly in line with my own, plus I liked pacman. I didn't read much about the Arch philosophy, "lightweight", "i686-optimized" and "KISS" were enough for me.

Now, laden with the wisdom of a few more years, I can safely say that I don't believe in philosophies and manifestos, they're like opinions of which everyone out there has got one. Neither do I believe that anyone was ever really bound by them. See what happened to Arch after Judd left; KISS went down the toilet or, worse, used to justify incredibly stupid decisions like the move of all binaries to /usr/bin. The adoption of systemd, however, wasn't a stupid decision - it was politics. Also, don't forget that systemd wasn't back then the monstrous octopus it has become these days so it wasn't that bad.

But, eventually, it became unbearable. What could I do then as a simple user? Cry, demand, request? No, I contributed. Thanks to the Gentoo devs, openrc and eudev came to be and thanks to Artoo and Aaditya (who had already left Arch for Manjaro), the first implementations of an alternative init system (that's how bad it had become, real init systems became alternatives) was possible and I glued some packages together to form an unofficial repository, [openrc-eudev], hosted on Dropbox no less. Then decided to launch a site about the project, ostensibly hoping to bring any leftover oldtimers together and secretly hoping for some help. A year or so later, Chris (Cromer) appeared and gave new breath to the project, now [arch-openrc], taking it over and overhauling it. The community Manjaro OpenRC edition from Artoo had also hit the street and I felt, for the first time in years, hope. In fact, they both were doing such a good job that I had semi-retired, tending only occasionally to the site. I thought I had done my part in saving the world.

Then, in July 2017, an email from Chris pulled me out of my nirvana: "We can't run forever behind Arch, we're forking it! Are you in?"

The rest is history. I won't retire any time soon. Yes, fuck philosophies; we've got work to do!

P.S. We do care about our users, we do get frustrated about reported installation errors, we do try to sneak out of our real life jobs to fix bugs. We would certainly like to satisfy every single reasonable request, implement more init systems, have thousands more packages in our repos but we just can't at the moment as our resources are limited. Want a philosophy? Here's one: join us!

[EDIT: systemd-free.org is now systemd-free.artixlinux.org]

Re: Artix "philosophy" in terms of the system

Reply #16
Thanks all, though as pointed out the "philosophy" seems to be "we package and develop init agnostic Artix distribution" which is fine.  I agree that there is no "we" (outside the dev team and even then it's more a common purpose rather than a guided vision).  I just dont want to be here a year or more later and find out that dev1 and dev2 who put the most work into it decide that next-big-iteration-of-package is the ONLY way to go.  As long as there is freedom (and room) for choice, I'm fine with whatever you guys do.

all I ask is not to be forced into any particular box  ;)

Re: Artix "philosophy" in terms of the system

Reply #17

all I ask is not to be forced into any particular box  ;)

Hmm, that is somewhat difficult.
Just an example, we already forced openrc to be default init.
While packages are compiled to be no systemd, ie supporting generally any init except systemd, we got the openrc layer on top.
The artix team could decide in the future to switch to s6 to be default init.
Other example is elogind, its default and not consolekit2 for technical reasons.
Point is, "force" is basically a two edged sword, we also do it in fact.

Re: Artix "philosophy" in terms of the system

Reply #18
Hmm, that is somewhat difficult.
Just an example, we already forced openrc to be default init.
While packages are compiled to be no systemd, ie supporting generally any init except systemd, we got the openrc layer on top.
The artix team could decide in the future to switch to s6 to be default init.
Other example is elogind, its default and not consolekit2 for technical reasons.
Point is, "force" is basically a two edged sword, we also do it in fact.
Why not include consolekit2 in the system repository for those that would prefer it over elogind? Is it simply because of the trouble-shooting that could come from it?

Re: Artix "philosophy" in terms of the system

Reply #19
Why not include consolekit2 in the system repository for those that would prefer it over elogind? Is it simply because of the trouble-shooting that could come from it?

The state of ck2 atm doesn't justify the effort. It would require many separate packages specifically compiled against ck2, eg pam, polkit, NM, sddm, plus a kernel change audit_syscall enabled, and  unmaintained pm-utils for power management and unmaintained cgmanager for cgroups1. Most DEs work much better with elogind, and with ck2, you also need specific ck2 patched DE packages.
Not gonna happen any time soon to see ck2 in repos, unless ck2 closes the gap to elogind and reaches feature parity.

Re: Artix "philosophy" in terms of the system

Reply #20
For those at home keeping score...
 Eric S. Raymond's musings about the "Cathedral and the Bazaar" have been turned into the "Cathedral and the Big Box Stores" thanks to the likes of Red Hat, Gnome, et al. Seems that TRUE user contributed based software is largely disappearing. And its because of this "branding", "rapid release", and agile methodology pushing for linux distributions from RH, Gnome, et al. that gave rise to and forced systemd  on us all. (IMHO)

So when the systemd fiasco finally hit Arch - I was bummed. I chose not to have that software. Like nous stated above, gentoo (or the funtoo derivative headed by former gentoo BDFL Daniel Robbins) was too much on compiling from source. My server is setup with funtoo, but I wanted something for the desktop I could "just use", albeit without systemd. Choices were/are few and far between. At least Google was sympathetic and pointed me in useful direction towards the earlier incarnation of nosystemd Arch.

And here we are...
What is the Artix philosophy? Arch without the crapware?
Or am I putting too blunt a point on this...

Re: Artix "philosophy" in terms of the system

Reply #21
And here we are...
What is the Artix philosophy? Arch without the crapware?

Even if only for keeping systemd out, yes.

And, personally, I'm all for unofficial user repos; for example one for consolekit2 along with the assorted packages @artoo mentioned, would be the perfect example of user contribution.
It would provide a ready-to-use testbed that would show the status of ck2 and in due time, when/if ck2 implements missing features, receives support from DEs and works well enough for everyday use, it could be moved to the official repos with a vengeance. Devs love ready meals!

Re: Artix "philosophy" in terms of the system

Reply #22
For those at home keeping score...
 Eric S. Raymond's musings about the "Cathedral and the Bazaar" have been turned into the "Cathedral and the Big Box Stores" thanks to the likes of Red Hat, Gnome, et al. Seems that TRUE user contributed based software is largely disappearing.


no.  that was never the model anyway

Re: Artix "philosophy" in terms of the system

Reply #23


How many users used and loved the Arch BSD scripts simple to use,  just 1 file rc.conf handled the whole management of startup and startup apps so simple that is what drew me to arch.

The good of Arch still  is its stability forget the crap Manjaro users are fed.
The bad is forcing users to use systemd instead of embracing choice, which is a back stab as  choice is what made Arch popular in the 1st place.

 The user chose every single item they installed now its mainly meta packages split down that in itself is still better than no choice at all as per the Manjaro, you will choose a desktop we will choose what you use on it by default.

Re: Artix "philosophy" in terms of the system

Reply #24
The polarization between individual pkg developers and "corporate" structures I believe is very misleading and false.  And I am trully devoted in supporting the underdog against the powerful even with neighborhood cats.
Not all collective effort has to be of the corporate hierarchical type.  The true separation, I believe, exists between formal organization and loose "free" association, to which also individual or small group enterprise belongs.  The ten people that make up the team named Urnix (imaginative name) are all really nice, friendly, helpful, tireless, good friends, ..... but Urnix is just a name for that small group of friends.  Urnix stands for nothing, no defined goal, no publicly known organizational structure, we don't know what they believe in or why they are doing it.  On the other hand you may have three who are not even friendly between them or to anyone else, under the nametag Strunix that has very strict commitment to specific goals, the way they work, committed to why they are doing what they are doing, and what they will not do to achieve their goal.  I go with Strunix, Urnix or Joeblo are not to be trusted or even worth a try.

The real difference is expected commitment to continuity.  An individual or a loose/free association at any given point vanishes in thin air.  A formal organization with well defined goals, values, principles, commitment, has continuity unless all long term goals have been met.  Even if after a while not a single person from the founding members is around (quit, died, thrown out, etc.) the organization still exists as long as the goals of its foundation have not yet been met.

Fungalnet can publish the world\s best browser tomorrow, that is most secure, light in resources, has zero bugs, and more functionality than any other.  Who is to say that it will be around and functional when the system it runs on has changed, the day after tomorrow?   Then fungalnet vanishes and reappears as JackO and publishes a great mail/news reader pkg?

The hell with the "little guy" then!  It is not the size that matters but the content of what it is they stand for.


Re: Artix "philosophy" in terms of the system

Reply #25
Arch without systemd is good enough for me!  8)

Have been using archbang artix/openrc for a few weeks now and loving it. Well done to the devs on making simple init scripts like openrc do there best with arch linux.

Re: Artix "philosophy" in terms of the system

Reply #26
Roy Maples?

Re: Artix "philosophy" in terms of the system

Reply #27
Philosophy is everything. It governs every facet of our lives, yet modern society is largely ignorant of that fact.
Same goes for this question.
In my opininin the reason the linux kernel project has been so outstanding,  consistently,  for so many years is a clear,  often stated,  base philosophy.  A philosophy brutally enforced by LT and that is a factor in almost every code decision.
So here is my attempt at a Artix Flag Manifesto. ( in order of priority)
1)  Bang for Buck.
Concentrate effort in places that give the greatest benefit.
2) Upstream is not God.
Upstream GNU projects can AND DO get it wrong, it is up to the only guys left in the chain , the distro, to correct those mistakes. Likewise it is up to the distros to pressure upstream to head in the right direction.
3) End users do exist.
Why do we do this if not to , in some way, satisfy an end user? Without a user Linux is an excercise in computer science, just like Unix was. Its analogous to a chef,  you have to make food that most people like, but you CAN NOT make food that everyone will like.
4) Linux is not alone.
Ms, Apple, Android, BSD,  do exist, plus a host of other wierder Linux based projects. No reason to put the blinkers on. And no reason not to directly compete.
5) Everyone contributes.
Only really big distros can afford apathetic users, in these early days everyone should contribute, in whatever way they are capable.

Re: Artix "philosophy" in terms of the system

Reply #28
Quote
Philosophy is everything. It governs every facet of our lives, yet modern society is largely ignorant of that fact

pontification?

Re: Artix "philosophy" in terms of the system

Reply #29
pontification?
I have no idea to tell you the truth
Maybe?
But then are we not into a English language discourse?
Am I being informative or pontificating. ? Does it reallly matter?
But if it really really offends will re-write ( honest gov ) lol.