Your words made me look again into the documentation:

If your APT configuration mentions additional sources besides bookworm, or if you have installed packages from other releases or from third parties, then to ensure a reliable upgrade process you may wish to begin by removing these complicating factors.

I hadn’t realized that “removing these complicating factors” meant removing these packages, not just disabling their repositories. The wording is terribly vague.

Now I don’t say anything against your experience and the conclusions it has led you to.

But my experience was that only repositories were automatically disabled and packages stayed in their place. The upgrades went through smoothly, things did not break. Were I forced to uninstall these packages and look for their replacements afterwards, I’d be quite annoyed. Maybe not as much as you, when you were forced to reinstall the system.

I’ll conclude for myself that both paths can lead to happy outcomes as well as to poor outcomes. Thank you for sharing!

Well, people do not follow instructions and their systems get broken 🤷 To a much larger extent than an orphaned package

I had smooth sailing with Ubuntu for many years, but I don’t judge other people’s choices

Wow, what a catch! I should really pay more attention to people… Okay, this person having a really terrible upgrade would perfectly explain what I’m seeing.

Here’s a gold star for winning this topic 🌟

Is this worse than an upgrade which breaks the system?

What do you think of the observation that Linux@lemmy.ml had a surprising amount of reports of bungled upgrades?

Oh, that’s bad luck… Sorry to hear that!

Yes, the change of the Lemmy user base was on my list of possible causes

Thank you, I see what you mean. I think there’s a flaw in this logic, but I would rather not dive deeper into this topic.

they release unreleased and/or unstable software

Is this true even for the point LTS releases?

The first two pages of Linux@lemmy.ml show me 6 entries about such upgrades 🤷 Do you subscribe to this community?

Thank you! This selection bias was one of the possibilities I’ve considered…

Thanks to differences in personalities and personal circumstances, there are usually enough people on the spectrum of risk aversion to test the pre-releases

Just the first page of linux@lemmy.ml shows me three Debian 13 related posts beside mine

Thank you for answering the actual question! This is a new angle I haven’t considered before.

This is a wise strategy used by patient people!

Do you think the number of less wise Debian users has changed in recent years?

What other hoops should I have jumped through to avoid being perceived as “generate shade”? I thought I did my best.

Do you think servers have it easy compared to personal computers? No pesky audio/video cards, bluetooth, etc…

Spoken with true wisdom!

Do you think the number of less wise Debian users has changed in recent years?

Happy to hear that!

But what is your opinion about the amount of complaints you see recently?