[Discussion] Why are severely broken point releases being released for Bookworm

[pizza-rat]

This thread is a demonstration of...

... a very bizarre over-defensiveness and strawmanning from a couple of Debian oldheads and enough diversionary nonsense that you'd think this was a political debate

[steve_v]

It is doing the project a huge favour, because today's Debian users are tomorrow's Debian developers.

A fine argument, until you set it alongside:

Your assumption that everyone wants to become an expert with computers is just as wrong as the assumption that everyone who drives a car wants to be able to repair their own brakes and change the engine.

Which is it then? New users who freak out if the GUI goes poof shouldn't be expected to learn how to fix it or use a CLI, or those users are the next developers? Can't really be both now, can it?
Selecting a different kernel from the bootloader isn't "changing the engine" anyway, not by a long shot. It's changing a tire (which every driver should know how to do) or swapping seat covers.

You shouldn't be as naive as IBM here, as IBM was with OS/2 and because of this, among other things, lost the OS war against Microsoft.
Microsoft took care of gamers, first with WinG, then with DirectX and today they benefit from those who became developers for Windows software.

The dealings between IBM and Microsoft at that time were far more complicated than "Microsoft catered to gamers", as you are doubtless well aware. That's a very small part of the picture.

a very bizarre over-defensiveness and strawmanning from a couple of Debian oldheads and enough diversionary nonsense that you'd think this was a political debate

It kinda is, or at least a philosophical one. GNU/Linux (and FOSS in general) has always been, at least in part, about the freedom to get your hands dirty and removing the distinction between user and developer.
Claiming users shouldn't need or want to know how to tinker with the system is claiming they'd willingly throw out freedom #2, as is becoming reliant on non-free software to the point of considering it a "critical" package and/or being "lost" without it.
The entire free software ecosystem was created by hackers, for hackers, and for those of us who jumped in the deep end and figured out how to swim, not even wanting to learn is frankly quite bizarre... As is using freee software seemingly without any desire to excercise the freedoms it grants.

You say "Users shouldn't have to deal with this, they need a GUI to do anything.", I say "What an excellent opportunity to screw around, break stuff, and learn something new (e.g. installing a new kernel, using the machine without a GUI, or installing an alternative driver)." Disk space is cheap enough these days that backups are damn near free after all, what's the worst that could happen?
Similarly, you say "Critical component", I say "One driver among many, interchangeable and not even strictly required to use the system."

Whichever way you look at it though, the fact is that solving the current nvidia SNAFU is but a boot menu selection or 'apt install' away. Why it's apparently such a big deal this week I have no clue, why not just deal with it and move on...
Or perhaps it's more productive to complain how broken everything is, and imply that somebody else should go ask the people building releases what's going on, because your machine is totally unusable without $proprietary_thing?

[M22]

Can't help but chime in again, when I saw "newcomer" repeated a few times here :

I am new to Debian, but not new to Linux. My first foray into Linux was Puppy Linux. From there I moved to Bodhi, TinyCore and was using Mint for a long time, until I settled down with Arch Linux for the past 5 or 6 years. I was comfortable with Mint but wanted to try Arch, and I got hooked to it, right from the command line installation. Surprisingly I learnt a lot from Reddit rather than the Arch forum, due to .... never mind. Anyway, I was okay with it until the rolling release mechanisms started to cause extreme exhaustion seeing that I need to fix something after an update. And though it rarely produced kernel panic, there were numerous other annoyances, making the updates a tiresome experience. And continuous updates which runs into hundreds of MBs everyday, even when checking in the evening after morning update, still runs into a sizeable amount. And using a "rolling" distro also gives you the feeling that you're never going to get to use a "fully baked" product. It's like "testing to eat , while you cook kind" of thing.

But time to move on, and I just want to use an OS with a piece of mind, with a "stable" distro. If required, just a touch of updates here and there to polish the release and that's it.

I knew about how stable Debian was from during the days I was exploring the various "flavors" of Linux. But somehow, the non-inclusion of non-free firmware in the default package, and the need to do the workaround or get the non-free firmware included ISO made me delay the installation. Almost all systems we use have one or another required "non-free firmware".

Finally, after getting exhausted with Arch, both, of updating and fixing, I decided to "settle down". And I looked back at Debian, mostly because of its reputation of being "rock stable", believe it or not, for those in the outside world, Debian means Stable. At least, the stable branch. And while downloading the ISO, I was surprised to learn Bookworm now includes non-free firmware too, in the default distribution. Installation was a breeze. And everything was perfect, true to it's "rock solid" reputation.

But, barely a month passed, I saw the kernel panic after a routine apt update and apt upgrade. Fortunately I found out soon from the forum what the cause was and on how to fix it, quick enough.

So, coming back to the newcomer, I'm new to Debian but not Linux. Imagine how I felt when I saw the kernel panic, after just a month of started using it, and reminded me of the dreaded days of seeing Arch Linux's release notes before updating. I have faced many annoyances when using Arch, but kernel panic....I can't remember if I faced it...other than a vague memory of something broke the GRUB. I expected problems in Arch, but never expected this to happen in Debian, that too after a very short time I started using it.

I also opened a thread recently here in the Discussion section.

As some people here have said, as a Stable branch OS, a problem, especially a one that can cause a kernel panic, seems like a big oversight, that should not have happened, more so in the Stable branch. I'm sure every update here usually had the reputation of being extensively tested before release, and if there is a known defect, it should be rectified before release, and there's no need to push it immediately.

That is the basis of a "Stable" branch, no?









P.S. All said, I like Debian and I guess this will be my final OS that I will depend on, unless something really goes wrong somewhere. (Hope not - I'm getting old, and I just need something that I can lean on safely.) It's fast, stable (ahemm) and works even better than Arch, from my experience. Even games run better in Debian. Compiling from sources in Debian is a much comfortable process compared to other distros I have used. Heck, even Bluetooth works much much better here, than in Arch.

And most importantly, the community here is very helpful when it comes to solving the problem, and they are helpful. (Of course, discussions sections will always have some bickering here and some shouts there, in any forum. But when it comes to giving help for questions and problems, I find this community is very nice. Thank you.

[Hallvor]

I have faced many annoyances when using Arch, but kernel panic....I can't remember if I faced it...other than a vague memory of something broke the GRUB. I expected problems in Arch, but never expected this to happen in Debian, that too after a very short time I started using it.

I'm sorry. As mentioned, it is a good idea to check the release notes before upgrading. It would also be advisable to avoid purchasing hardware from the company that Linus Torvalds has labeled as "the single worst" he has dealt with. Use the search function above to observe the multitude of issues associated with such hardware. (All-Intel hardware should be more or less trouble-free.)

https://www.youtube.com/watch?v=i2lhwb_OckQ

[alienspy]

I'm sorry. As mentioned, it is a good idea to check the release notes before upgrading.

Where the one should check it? On the web i have found only this - https://www.debian.org/News/2024/20240210. And it doesn't have much details. If it must be found in a mailing list, then in which one exactly? Thanks.

[pwzhangzz]

It would also be advisable to avoid purchasing hardware from the company that Linus Torvalds has labeled as "the single worst" he has dealt with. Use the search function above to observe the multitude of issues associated with such hardware. (All-Intel hardware should be more or less trouble-free.)

https://www.youtube.com/watch?v=i2lhwb_OckQ

Couldn't agree more.

Additionally, according to a recent report, the fourth generation RDNA graphic processors (RDNA4) are not supposed to be coming out until Q3 2024 but Linux kernel developers are already busy preparing RDNA4 drivers for the future Linux 6.9 release. Please see:

https://www.phoronix.com/news/Linux-6.9 ... -New-Stuff

Comparing to Nvidia chips, for which Linux kernel developers have to do extra work--sometimes incur the risk of being unabashedly chastised/trashed for doing an impossible job (as "beautifully" demonstrated in this thread)--to retrospectively reverse engineer many of their unpublished specs, the open-sourced amdgpu driver should be the model for future hardware development. I fully support this model with both my mouth and my pocket book (and my feet). Of course, free software also means that Linux users have the freedom to choose their own hardware. But just don't trash Debian developers for not making enough sacrifices. The title of this thread is not only annoying, but nauseating, demeaning, and disgusting.

Also, contrasting the specs-withheld nvidia chips, the prospectively developed amdgpu drivers also mean that Linux kernel developers' feedbacks will greatly aid hardware development.

[Borg]

Comparing to Nvidia chips, for which Linux kernel developers have to do extra work--sometimes incur the risk of being unabashedly chastised/trashed for doing an impossible job (as "beautifully" demonstrated in this thread)-

Can you explicitly support your claims with evidence?
Which comment exactly in this thread do you mean?

The title of this thread is not only annoying, but nauseating, demeaning, and disgusting.

I have to completely disagree with you, the title doesn't even begin to reflect what you're accusing the title of here.
The title essentially consists of just a legitimate question, which was then clarified in the first posting.

[Borg]

It is doing the project a huge favour, because today's Debian users are tomorrow's Debian developers.

A fine argument, until you set it alongside:

Your assumption that everyone wants to become an expert with computers is just as wrong as the assumption that everyone who drives a car wants to be able to repair their own brakes and change the engine.

Which is it then? New users who freak out if the GUI goes poof shouldn't be expected to learn how to fix it or use a CLI,

Your question is nonsensical, let me illustrate this using the following analogy:
Do you seriously expect a young baby to be able to walk, talk and do math problems on the first day after birth?

And that's how it is with a new operating system. The new users use it first because they want to run software. For example they want to play games. And you can't expect anything more from them at first.
Most people will continue to use an operating system only to do their actual tasks, but there will also be those who then want to know more about the operating system itself and learn how it works, who then want to know how software works and how to write it. Then they learn programming etc. But I already told you all that in my post to which you replied.

Selecting a different kernel from the bootloader isn't "changing the engine" anyway, not by a long shot.

Nobody said that, it wasn't about loading the previous kernel. but changing the driver does it, like some people suggested here in some threads to this problem. Especially for those who are lost without a GUI.

It's changing a tire (which every driver should know how to do) or swapping seat covers.

Changing a tire is not like changing a completely different driver. A completely different driver will result in completely different behavior and there is a risk that it will not work. The danger is even greater for beginners who can't get any further without a GUI because they can no longer help themselves. So it's like if you replace the braking system and then the brakes are too weak to stop the car in time.

You shouldn't be as naive as IBM here, as IBM was with OS/2 and because of this, among other things, lost the OS war against Microsoft.
Microsoft took care of gamers, first with WinG, then with DirectX and today they benefit from those who became developers for Windows software.

The dealings between IBM and Microsoft at that time were far more complicated than "Microsoft catered to gamers", as you are doubtless well aware. #

Read more carefully, i said the words: "among other things".

That's a very small part of the picture.

It was definitely not a small part.
Four reasons are generally responsible for the defeat of IBM and its OS/2, in this order:
1. IBM shipped OS/2 with a Windows 3.x license (Red Spine box package). This meant that the developers saw no reason to develop for OS/2, because if they only developed for Windows, they could cover both markets thanks to the included Windows license in OS/2. Doing extra work for a OS/2 version wasn't necessary.
2. IBM didn't care about good hardware support.
3. IBM didn't care about the game developers, who practically decide for the entire home market which OS will be installed there.
4. IBM spent unnecessary money on football perimeter advertising, something that no one needs and that doesn't benefit an operating system at all. They should have used the money to pay the hardware manufacturers to hire programmers to develop drivers for their old hardware. That would have done a lot more. But that's probably what happens when you have the wrong people in the marketing department who don't know anything about computers and think that sporting events are somehow important.

[dilberts_left_nut]

Enough.
The problem is solved and reckons have been had.
Move on.
Thread locked