[SOLVED] Using testing repos

[Member dirkd]

I did a fresh install of Antix19.5 and I wanted to switch to the testing repos. I *think* I did it as recommended in previous posts on this topic: commented out all previous active (stable) repos, and uncommented two lines in debian.list and antix.list mentioning the testing repos. Then

sudo apt-get update
sudo apt-get full-upgrade

However, this breaks the system completely. The upgrade command ends with dependency errors, among others with libcrypt.so (if I remember it rightly) and libc64. If I try another sudo command I get the question if my account is locked. As it stands, I cannot even logout or shut down the system, and after rebooting I can’t login.

So what is the correct procedure for switching to ‘testing’?

• This topic was modified 9 months, 3 weeks ago by Brian Masinick.

[Forum Admin anticapitalista]

Helpful

Up

0

::

Best to start with antiX-21 not antiX-19 series

Philosophers have interpreted the world in many ways; the point is to change it.

antiX with runit - leaner and meaner.

[Member dirkd]

Helpful

Up

0

::

I did use Antix21 for a while, but (as discussed in previous posts), for unexplained reasons it cannot boot on my present hardware. And I really tried my best to fix it, but to no avail. So I had to revert to a working Antix17 system that I had (luckily) still on my hard drives. But now that the Antix17 repo has been retired, it’s either Antix19 or something completely different.

I found some log files produced during the unsuccessful full-upgrade.

Preparing to unpack …/13-libgcc-s1_12.1.0-2_amd64.deb …
Unpacking libgcc-s1:amd64 (12.1.0-2) …
Replacing files in old package libgcc1:amd64 (1:8.3.0-6) …
Setting up libgcc-s1:amd64 (12.1.0-2) …
Setting up libc6:amd64 (2.33-7) …
/usr/bin/perl: error while loading shared libraries: libcrypt.so.1: cannot open shared object file: No such file or directory
dpkg: error processing package libc6:amd64 (–configure):
installed libc6:amd64 package post-installation script subprocess returned error exit status 127
dpkg: dependency problems prevent configuration of libblkid1:amd64:
libblkid1:amd64 depends on libc6 (>= 2.33); however:
Package libc6:amd64 is not configured yet.

dpkg: error processing package libblkid1:amd64 (–configure):
dependency problems – leaving unconfigured
dpkg: dependency problems prevent configuration of libext2fs2:amd64:
libext2fs2:amd64 depends on libc6 (>= 2.33); however:
Package libc6:amd64 is not configured yet.

dpkg: error processing package libext2fs2:amd64 (–configure):
dependency problems – leaving unconfigured
Errors were encountered while processing:
libc6:amd64
libblkid1:amd64
libext2fs2:amd64
Log ended: 2022-07-13 12:04:32

My debian.list

# Note:If you want maximum stability, keep to the default Stable/buster repos.
# Buster is the new stable release and will only recieve security
# updates, while Testing will be ‘rolling’.

# Debian Stable. Default for antiX-19.
# Use for maximum stability INSTEAD of the ‘rolling’ TESTING release concept.
#deb http://ftp.be.debian.org/debian/ buster main contrib non-free
#deb http://security.debian.org/ buster/updates main contrib non-free
#deb-src http://ftp.be.debian.org/debian/ buster main contrib non-free

# Debian Testing.
# Testing enabled for ‘rolling’ release.
deb http://ftp.be.debian.org/debian/ testing main contrib non-free
deb http://security.debian.org testing-security main contrib non-free
#deb-src http://ftp.be.debian.org/debian/ testing main contrib non-free

# Multimedia Stable and Testing
# Use to install libdvdcss2 and codecs.
#deb http://www.deb-multimedia.org buster main non-free
#deb http://www.deb-multimedia.org testing main non-free

###### Debian Unstable/Sid##########
###### Use at your own risk! ########
#deb http://ftp.be.debian.org/debian/ unstable main contrib non-free
#deb http://www.deb-multimedia.org sid main non-free

My antix.list

# Use with Debian Stable/buster repositories. Set as default for antiX-19.
#deb http://mirror.tiguinet.net/mx/packages/antix/buster buster main nonfree nosystemd
#deb-src http://mirror.tiguinet.net/mx/packages/antix/buster buster main nonfree nosystemd

# Use with Debian Testing/’rolling’ repositories.
deb http://mirror.tiguinet.net/mx/packages/antix/testing testing main nonfree nosystemd
#deb-src http://mirror.tiguinet.net/mx/packages/antix/testing testing main nonfree nosystemd

# Use with Debian Sid repositories.
#deb http://mirror.tiguinet.net/mx/packages/antix/sid sid main nonfree nosystemd
#deb-src http://mirror.tiguinet.net/mx/packages/antix/sid sid main nonfree nosystemd

All other sources contain nothing but commented lines.

• This reply was modified 9 months, 3 weeks ago by dirkd.
• This reply was modified 9 months, 3 weeks ago by dirkd.

[Forum Admin anticapitalista]

Helpful

Up

0

::

You could try upgrading in stages.

1. With original sources, do a dist-upgrade and see if it still works ok.
2. Change to antiX-21 sources and then do a dist-upgrade and see if it still works ok.
3. Now, change sources to testing, do a dist-upgrade and see if it still works ok.

Philosophers have interpreted the world in many ways; the point is to change it.

antiX with runit - leaner and meaner.

[Member sybok]

Helpful

Up

0

::

Hi, as a last attempt, you may try to (repeatedly) run ‘sudo apt install –fix-broken’.
This may correct the not-yet set library and help you get out of the dependency hell.
If there are any packages suggested to be removed (which you expect to be important and not to be replaced in the update), please copy the list into a TXT file for possible recovery/reinstall if this approach leads to a system that boots and allows you to login (at least using a console mode).

As @anticapitalista pointed out, upgrade/update from antiX-19 (based on Debian 10) to antiX-testing (already ahead of antiX-21, based on Debian 11) is likely a leap too big to be successful.

[Moderator caprea]

Helpful

Up

0

::

Also , sometimes a sudo apt autoremove is needed in my experience
when I saw these errors

[Member dirkd]

Helpful

Up

0

::

@anticapitalista:

I never thought it would work, but it seems I’m almost there. I *seem* to have a running system with bullseye repos and three kernels 4.9 – 5.10.88 – 5.10.104.

Now I need one of the 5.10 kernels to support the built-in wifi6 interface, but these mess up the graphical login. The 4.9 kernel brings me to the desktop flawlessly, but doesn’t support the wifi. I hypothesised that maybe I needed slimski in stead of slim, so I installed that too. Then I booted a 5.10 kernel, Alt-Ctl-F1 to the text-console, and issued the command

sudo slimski

That worked ONCE to bring me to the graphical login screen. After rebooting, I see a black screen, while the console screen (Alt-Ctl-F1) informs me that slimski (or slim if I change the default displaymanager) has been started, although no graphical login is shown.

Any suggestions on how to repair the graphical login?

@sybok: thanks, maybe useful in the near future
@caprea: did that, as suggested by the system while doing the full-upgrades

• This reply was modified 9 months, 3 weeks ago by dirkd.
• This reply was modified 9 months, 3 weeks ago by dirkd.

[Member dirkd]

Helpful

Up

0

::

Some more diagnostics: in a whim I decided to uninstall slim. Although documentation suggests there’s nothing wrong having both slim and slimski, only one of them is needed. To my surprise, I now had – consistently – a working graphical login screen.

BUT:

This state of the system is only preserved as long as you do warm reboots. Once you shut down the system and power on again, the graphical login screen is no longer shown when using the 5.10 kernels. The only way I know, for now, to get it back is by booting the 4.9 kernel, re-installing slimski using synaptic, then doing a warm reboot.

Now, is there a clue to the problem there? I wonder if it is not some kind of timing problem. I notice the boot sequence on this new system is lightning fast, even to Antix standards. Much faster it seems then what I see using my Antix17 installation on the same system. If so, what can be done?

EDIT: it seems unnecessary to actually reinstall slimski. Booting the 4.9 kernel after a power off/power on cycle, then doing a warm reboot seems enough to get a graphical login with a 5.10 kernel

• This reply was modified 9 months, 3 weeks ago by dirkd.

[Member dirkd]

Helpful

Up

0

::

It seems the system has cured itself. Don’t know how exactly: all I did is a series of warm and cold reboots for testing purposes. But for some time now I consistently get to the graphical login screen with a 5.10 kernel. So this can be considered solved. I now have bullseye system, starting from Antix19.5. Bullseye offers version 5 of the wine package in the stable repo, which was the main reason I wanted to use testing repo in Antix19.

Only problem remaining is that the dhcp client can’t connect to my router: it complains the network is down, which definitely is not the case. But I’m confident I will be able to fix that.

[Member sybok]

Helpful

Up

0

::

Glad to read you manage to upgrade and good luck with the DHCP client.

[Moderator Brian Masinick]

Helpful

Up

0

::

Following the suggestions (which you did), the autoremove can clean unnecessary packages and it’s possible to actually fix issues.
(repeatedly) run ‘sudo apt install –fix-broken’. (I do the equivalent, sudo apt-get -f install), particularly combined with the autoremove, clean, and autoclean keywords are also common ways to clean up similar messes. Any or all of these that you used, plus updating from a recent source all contributed to the solution. I’ll mark it solved; thanks!

--
Brian Masinick

[Author]

Posts