Skip to main content ↑ ↓
Topic: runit-artix-20180414-9  (Read 2691 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

runit-artix-20180414-9

Quote from: konimex – on
ATTENTION TO ALL RUNIT USERS:

See announcement

Since this thread has served its purpose, I'm going to close this thread by the time runit-artix-20180414-9 lands in [system].

Good thing you did, but I don't see the new one I start this one.
My system after all this turmoil and after some fine tuning and figuring out things has been running very well for days.
Suddenly I wake up,  I open a terminal and do an upgrade.
In terror I see a new runit-artix****-9  (5 versions past the previous one), it upgrades "no warnings or notices for the installation" (it lags a bit too long) but all clean.

Then I reboot

I end up in emergency shell

/sbin/init is missing it says!

Need I say more?

Re: runit-artix-20180414-9

Reply #2 –
Also: 
This part of the wiki will no longer be needed once this transition has reached the stable repository

Code: [Select]
When you’re restarting, to use runit as your init system, make sure you add a kernel parameter init=runit-init in your desired bootloader. For example, in GRUB:

linux /boot/vmlinuz-linux root=UUID=978e3e81-8048-4ae1-8a06-aa727458e8ff
init=/sbin/runit-init

or you if you want to set runit as your default init system, can symlink runit-init to /usr/bin/init instead.

As runit-init has been replaced by init, right?

I removed it and it starts

Re: runit-artix-20180414-9

Reply #3 –
I used the system all day yesterday, after the first wave of artix-sysvcompat ordeal and it run fine.  I do an upgrade last night and there is a new version:

Code: [Select]
gremlins/artix-sysvcompat 0.3.7-5 (base)
    Artix Linux's wrapper program for OpenRC and runit's init & shutdown scheme

I try to boot it now --> nothing --> nada --> sysvincompat

At least, least, can someone clue us in to what the *** is going on?
No wonder nobody talks on this forum again, if artix is the only thing they have they have been silenced by pacman

Code: [Select]
with    artix-sysvcompat-0.3.7-3-x86_64.pkg.tar.xz
# ls -lh /sbin/init*
lrwxrwxrwx 1 root root 7 Apr 28 00:12 /usr/bin/init -> default

with    artix-sysvcompat-0.3.7-5-x86_64.pkg.tar.xz
lrwxrwxrwx 1 root root   11 Apr 28 14:07 /usr/bin/init -> openrc-init
on a system running Runit!

So this package artix-sysvcompat was made yesterday for OpenRC only systems?

At least pick 2-3 of us volunteers and email us the pkg to be tested, alarm us of the possibilities, place it in staging, but don't release it without any notes of what is going on.
As far as we know here there are 2 devs working on runit and 2 users

Re: runit-artix-20180414-9

Reply #4 –
Quote from: fungalnet – on
I used the system all day yesterday, after the first wave of artix-sysvcompat ordeal and it run fine.  I do an upgrade last night and there is a new version:

Code: [Select]
gremlins/artix-sysvcompat 0.3.7-5 (base)
    Artix Linux's wrapper program for OpenRC and runit's init & shutdown scheme

I try to boot it now --> nothing --> nada --> sysvincompat

At least, least, can someone clue us in to what the *** is going on?
No wonder nobody talks on this forum again, if artix is the only thing they have they have been silenced by pacman

Code: [Select]
with    artix-sysvcompat-0.3.7-3-x86_64.pkg.tar.xz
# ls -lh /sbin/init*
lrwxrwxrwx 1 root root 7 Apr 28 00:12 /usr/bin/init -> default

with    artix-sysvcompat-0.3.7-5-x86_64.pkg.tar.xz
lrwxrwxrwx 1 root root   11 Apr 28 14:07 /usr/bin/init -> openrc-init
on a system running Runit!

So this package artix-sysvcompat was made yesterday for OpenRC only systems?

At least pick 2-3 of us volunteers and email us the pkg to be tested, alarm us of the possibilities, place it in staging, but don't release it without any notes of what is going on.
As far as we know here there are 2 devs working on runit and 2 users


You know that's literally the purpose of [gremlins], don't you?

If you, or anyone else doesn't want their systems to break because of changes in [gremlins], then don't use [gremlins] at all. That's the reason [gremlins] exists, because we'll get feedback before we really roll it out in stable ([system]).

Artix is based on Arch, so we at Artix follow Arch's repo system.

[gremlins] is the equivalent of [testing] in Arch. On the other hand, the reason we don't encourage (nor tell people) to use [goblins] aka [staging], I quote, "is only meant for backend developer use".

Anyway. If you want to know whatever happened to artix-syscompat, since our repo is open-source, please see the PKGBUILD history here.
now only the dinit guy in artix

Re: runit-artix-20180414-9

Reply #5 –
Quote from: fungalnet – on
At least pick 2-3 of us volunteers and email us the pkg to be tested, alarm us of the possibilities, place it in staging, but don't release it without any notes of what is going on.
As far as we know here there are 2 devs working on runit and 2 users

Artix is a double-blind distro; some guinea pigs users know they're using testing and some don't. And then, there's the control group regular userbase that enjoys the bleeding edge stability of stable repos, thanks to the testing victims!  8)

Please, take the above as a joker's compliment. Both @artoo and @konimex have worked Sunday overtime to fix the issues you've been raising. Cheers!

Re: runit-artix-20180414-9

Reply #6 –
If you remember when the troublesome OpenRC 35 came out I was the first guinea-pig that volunteered to try out all the breakage, I don't mind it at all.  A little communication would help instead of smelling my finger nails (and toe nails on this one) trying to get answers.  I love breakage and fixeage as much as anyone, but not in the dark and underwater.

Re: runit-artix-20180414-9

Reply #7 –
Quote from: konimex – on
You know that's literally the purpose of [gremlins], don't you?

Instead of preaching to someone who has never run stable anything, there have been 2 versions of -compat (-5 and -6) today that have broken my runit system and although you manage to write all this there is no mention of why it is breaking -3 (as in not ever reaching stage2).  All I ask is a bit of a clue.  @artoo said there might be a possible fix in -6 and it breaks yet again.  So should I try -7 -8 -9 in the dark without instructions?  That's all!

The definition of "testing" is not an acceptance of "yes, we tried, we messed up, we are working on it, sorry" which is very human and very acceptable.
Save the preaching and clue us in what is going on.

Re: runit-artix-20180414-9

Reply #8 –
I think your tone is inappropriate.

Re: runit-artix-20180414-9

Reply #9 –
Quote from: artoo – on
I think your tone is inappropriate.
I agree....much is constantly changing w/gremlins, and very quickly indeed...especially as concerns runit init. I feel a good measure of patience should be employed. Realistically, breakages should be expected.

Best regards.
We should try to be kind to everyone.....we are all fighting some sort of battle.

Re: runit-artix-20180414-9

Reply #10 –
I said what I said, message got through, end of story
 

Re: runit-artix-20180414-9

Reply #11 –
Dude, gremlins is an unstable/testing repository. You should always double check PKGBUILDs before you blindly upgrade anything absolutely essential like init. If you don't want to expose yourself to possible stability issues (which is perfectly understandable),  then just stay on stable.

Re: runit-artix-20180414-9

Reply #12 –
O dear I feel their is a lot of bad feelings at this moment in time runit is a good init but it needs time to integrate with Artix , but their is no need to fall out with artoo or any member of the devs that is just defeatist to the whole project and just makes the project take backward steps we need to step forward at all times not fallout with each other.

Re: runit-artix-20180414-9

Reply #13 –
Quote from: Dudemanguy – on
You should always double check PKGBUILDs before you blindly upgrade anything

I have yet to see a pkgbuild say the pkg will not work and if you make the mistake of installing it you are on your own.
Ultrix was very unstable when it first came out, I remember that much.  Do you?

Re: runit-artix-20180414-9

Reply #14 –
You can certainly spot major changes in a package if you look at some history. You mentioned that you were worried about runit-artix suddenly jumping up 4 versions. Instead of just hitting "yes", why not take a look at the PKGBUILD history? Clearly, those aren't trivial changes and I wouldn't update until I was totally certain what going on.