Lubuntu update caused big problems (help needed)

Ubuntu
21

Okay, I thought you are using an Ubuntu dependency on Debian, when it’s actually the opposite.

So, you shouldn’t use a Debian package on Ubuntu and vice versa, as that might break things, as you have just experienced.

1 Like
22

Thank you I will try that

In the meantime I tried another system-update, setiing `base-files`
 and `snapd` on hold first.

After a reboot and a cold-start I can confirm it went well and there 
are complaints whatsoever.
So it seems to be **very** certain that either of the two
 aforementioned packages **has to be** the cuplrit.

Well done. I think you should wait a bit longer now, see if any bug info about base-files or snapd appears.
Then if nothing appears, backup and try updating base-files

Regards
Neville

23

Hi to you all again, :wave:

thanks a lot for your input.

@ahmetbombaci:

I´ve never heard of exa before and looked it up on exa: A Modern Replacement for the ls Command .
It looks very interesting indeed. I´ve been using the ll command until now and it serves me well.
Using the fish-shell it says:

functions ll
# Defined in /usr/share/fish/functions/ll.fish @ line 4
function ll --wraps=ls --description 'List contents of directory using long format'
    ls -lh $argv
end

so it seems to be an alias for ls -lh. But exa´s colourful output looks very nice indeed. :wink:
It´s a shame though you ran in trouble while using it.

@Akito:

thanks for pointing that out. :+1:

@nevj:

Hi Neville,

As far as script is concerned it says on bsdutils › Wiki › ubuntuusers.de :

The disadvantage of using script is that all special characters are also logged, so the output file is full of control characters and ANSI escape sequences.
However, the contents of the output file can be almost completely cleaned using the col program, which creates another file (cleanedfile in this example):

col -bx <typescript >cleanedfile

The pager less is suitable for reading the log files – especially the uncleaned ones; you can also use cat for very short files:

less -r LOGFILE

(translated from German)

Thanks. :heart:

I have to admit I haven´t waited for much longer and as a second attempt I installed base-files and snapd.
I gave in to an urge to try again. :blush:

I had timeshift make a snapshot before so that I wouldn´t lose the things I had freshly updated.

This time - believe it or not - everything went smoothly. :+1: :smiley:

No glitches whatsoever and none of the problems mentioned in my first post appeared.
I rebooted my system after the latest upgrade and also performed a cold-start. In both cases: no problems.

Phew, am I glad now. :slightly_smiling_face:

I still don´t know why the problems initially showed up.
The only differences between the first scenario and the scenario yesterday were:

  • the second time I updated in 2 steps, the first one excluded the base-files and snapd packages, the second one just updated them exclusively
  • in the meantime two additional packages were added to the update: libexpat1 and
    libsasl2-modules

Well, in the end it seems to have gone well and my system is up-to-date at the moment. :slightly_smiling_face:
For what it´s worth I once more learned a few things thanks to the kind help from all of you. :+1:

Thanks a lot. It´s very much appreciated. :heart:

Many greetings from Rosika :slightly_smiling_face:

1 Like
24

UPDATE / ADDITIONAL INFO:

Hi again, :wave:

I´ve subscribed to the rss-feed provided by Ubuntu security notices and just received an interesting message:

USN-5292-4: snapd regression

Hmm, that made me wonder, especially in view of the fact that the snapd-package was one of the two “candidates” that might have been responsible for the problem. :thinking:

The feed began thus:

USN-5292-1 fixed a vulnerability in snapd. Unfortunately that update introduced a regression that could break the fish shell

And that´s exactly the case with me. I´m using the fish-shell by default. :blush:

Taking a look at https://ubuntu.com/security/notices/USN-5292-4 and after that at
Bug #1961791 “2.54.3+21.10.1ubuntu0.1 broke Plasma Desktop when ...” : Bugs : snapd package : Ubuntu

the bug description began thus:

Yesterday I received snapd’s 2.54.3+21.10.1ubuntu0.1 update. And when I rebooted, my Plasma Desktop was broken. Applications that were still open worked (Firefox, Konsole), but task panel was absent, icons missing, … Basically the desktop session became unusable.

Note that I am using Fish shell as my default shell. When I use Bash however, the Plasma Desktop is working fine.

(bold by me)

Well there are some differences when comparing it to my setup:

I´m using

2.54.3+20.04.1ubuntu0.2 instead of 2.54.3+21.10.1ubuntu0.1 now

and the phenomena decribed were not quite the same with me , apart from the fact that it´s LXQt with me … but still:

I´m beginning to wonder if it could´ve been snapd instead of base-files which introduced the phenomena described in my first post :question: :thinking:

… especially in view of the fact that I use the fish-shell be default…

Curiously enough - and God-be-thanked - my latest update hasn´t produced any problems whatsoever. :smiley:

Perhaps it´ll remain a mystery forever. :slightly_smiling_face:

Well, I just wanted to let you know.

Many greetings from Rosika :slightly_smiling_face:

1 Like
25

Amazing analysis by @Rosika, as usual. I have nothing to add, you did great. :clap:

The only thing I could perhaps amend is the explanation of what a “regression bug” is.

A regression bug is a bug surfacing in previously working and perfectly fine implementations, just because a new functionality was added beside it. It’s hard to detect such bugs, because the previously working implementation was not touched. It just got broken as a side-effect from implementing a new and different feature, which seemingly is not related to the previously available one.

1 Like
26

Thanks so much @Akito for your praise.
I feel honoured indeed. :blush:

It´s great of you to provide the actual definition of “regression”.
I had a rough idea about it, but now I know it even better. :+1:

It´s really wonderul to be able to learn new things and enhance one´s knowledge in this fabulous forum.

You´re all highly educated and great experts. Very good teachers indeed. :+1:

Thanks a lot for putting so much time and effort into helping others. :heart:

Many greetings from Rosika :slightly_smiling_face:

1 Like
27

Akito that confuses me?
It says snapd regression
Does it mean snapd caused a bug in Rosika’s system, or does it mean they are going to regress snapd to an older version to avert a bug ?

And I agree with your sentiment. Rosika has done a marvellous job of diagnosis and remedy, and of keeping everyone informed.

1 Like
28

Hi Neville, :wave:

Thank you so much, Neville, for your compliment. :heart:
But I think the praise should go to you and @Akito and all the others that so diligently helped me in solving the problem. :+1:

I looked up the definintion of “regression” (software-related) here: Software regression - Wikipedia .

According to them:

A software regression is a type of software bug where a feature that has worked before stops working. This may happen after changes are applied to the software’s source code, including the addition of new features and bug fixes.

They may also be introduced by changes to the environment in which the software is running, such as system upgrades […]

So this. as I see it, proves @Akito to be very correct in his assessment:

In my case I may confirm that it had an impact on (the performance of, or at least the functionality of) my system.

I´m very glad the issue could be solved with the kind help of you nice people. :slightly_smiling_face:

Many thanks and have a great weekend.
Many greetings from Rosika :slightly_smiling_face:

1 Like