Debian User Forums

...42.

If that doesn't work, reboot and install Windows. Then your problems won't seem so bad. Plus, always remember that in some parallel universe your hardware is supported and you have Debian running perfectly!

Felt like a little humor today.

...maybe if more people invested in a healthy babelfish from a certified breeder, debian wouldn't seem so cryptic.

neuraleskimo wrote:If that doesn't work, reboot and install Windows

Corrupted Winblows registry
Win10 hangs after waking from sleep -> disable power saving mode
Forbes: Windows 10 users continue to suffer from problem after problem...

Indeed a good joke...

If there are infinite universes, then there has to be a Hellverse worse than our current one where installing Windows is indeed the answer.

stevepusser wrote:If there are infinite universes, then there has to be a Hellverse worse than our current one where installing Windows is indeed the answer.

LOL! +6.022x10^23

That's the one with MS Linux! Oh, wait that that might be the future after MS buys Canonical...uhm...crap!

neuraleskimo wrote:That's the one with MS Linux! Oh, wait that that might be the future after MS buys Canonical...uhm...crap!

MS Windows is pretty soon going to transform from "Windows subsystem for Linux" into MS Linux with a subsystem for Windows.

sickpig wrote:MS Windows is pretty soon going to transform from "Windows subsystem for Linux" into MS Linux with a subsystem for Windows.

That sounds like a joke but I did see it reported on Slashdot the other day. I have fears for Linux if MS becomes a major contributor to the kernel. They already contribute a small but significant amount of code. Undoubtedly someone will fork it so there's a Linux sans "Linux subsystem for Windows".

MS Azure has majority of Linux instances since ages. MS is submitting Linux kernel changes to better suit their Azure hypervisors to support ever increasing demand for Linux focussed HPC and general computing requirements.

MS is making heaps of money off of Linux. AWS and Google public clouds are heavily linux focussed. Google even uses their own Debian based OS internally.

I would be least surprised if in the comming years MS in the interest of making more money switch to linux and provide a portability hypervisor for windows based workloads.

MS mantra here is if you can't beat them join them.

edit - MS has publicly admitted Linux dominance in Azure in various articles

sunrat wrote:I have fears for Linux if MS becomes a major contributor to the kernel.

They were in the top five for kernel contributions for a few years running a while back and they are currently a Platinum member of the Linux Foundation.

Posted from OpenBSD

Head_on_a_Stick wrote:They were in the top five for kernel contributions for a few years running a while back and they are currently a Platinum member of the Linux Foundation.

That's true, but excluding improved support for MS hypervisors, they are mostly polluting the kernel with useless (and often just stupid) "extensions".
It looks like they are porting shitty solutions (interfaces) from Windows kernel to Linux kernel - so I suppose that Mr sickpig is right - they are heading to MS Linux.
A particularly good example is trampfd

Clearly Microsoft engineers are morons, who don't understand even the basics of system design - that's why they have "invented" trampfd, which basically delegates security measures from kernel to applications.
In no words I can express how stupid this solution is - but if someone wonders why there are at least 300 new viruses spawned for MS Winblows each single day, then the above may shed some light...

But even more worrying fact is, that Linux Foundation ignores what the members are doing in technical aspects, as long as they're paying for Platinum Memebership.

Where is Linus when we need him . . . a victim of cancel culture . . .

LE_746F6D617A7A69 wrote:so I suppose that Mr sickpig is right - they are heading to MS Linux.

In installments. First one being delivered soon -

That logo looks like a tsunami wave coming in . . . I guess that's the "extinguish" meme.

LE_746F6D617A7A69 wrote: That's true, but excluding improved support for MS hypervisors, they are mostly polluting the kernel with useless (and often just stupid) "extensions".

That's exactly the "fears for Linux" I mentioned above.
Embrace, extend, extinguish. It would apply to their incompetence just as well as historical commercial malfeasances.
Is it possible or too late to fork a non-MS Linux kernel? Or is BSD going to be the popular FOSS choice for those who care in the future? I'm certainly starting to consider it although I don't think it supports many of the Linux music production applications I currently use.

golinux wrote:That logo looks like a tsunami wave coming in . . . I guess that's the "extinguish" meme.

That's the plan. First port your major applications like vscode, edge and before you know it MS Linux is a living, breathing thing.

sunrat wrote:Is it possible or too late to fork a non-MS Linux kernel? Or is BSD going to be the popular FOSS choice for those who care in the future?

Technically, this is possible - but I doubt that any distro would dare to "take a lead" of kernel development ...
BSD kernels have a probem with SMP performance - it will take years for them to catch the Linux kernel in that aspect ...

Unfortunately I don't have a good answer...

The kernel configuration is really flexible. Can't any "features" you don't want just not be enabled?

LE_746F6D617A7A69 wrote:BSD kernels have a probem with SMP performance

DragonFly BSD fares pretty well for SMP support (although it's still not as good as Linux in that respect): https://www.dragonflybsd.org/performance/

And anyway performance isn't everything

Head_on_a_Stick wrote:

LE_746F6D617A7A69 wrote:BSD kernels have a probem with SMP performance

DragonFly BSD fares pretty well for SMP support (although it's still not as good as Linux in that respect): https://www.dragonflybsd.org/performance/

And anyway performance isn't everything

I disagree - the performance is the key factor - it defines the MIPS per Watt factor (thus the energy consumption/savings), but what is probably more important - the performance of SMP-dependant applications, which are becoming ubiquitous nowadays.

Perfect SMP scaling means the lowest possible energy/shortest time needed to perform an SMP tasks - and AFAIN BSD kernels can't assure that correlated (bound to a single process) SMP context switching will happen in a very short or nearly zero-time time window - that's the main problem (but I think that BSD developers could port the solutions used in Linux kernel - they are relatively simple).

I think that Linux kernel development slowly but inevitably changes the direction from a "robust code" to a "corporate code" - where the only counting aspects are the version numbers and the number of patches... (Quality Assurance is too expensive ...)

I'm seriously considering Haiku and BSD kernels a the future of Free, High Quality kernels, because those projects are not polluted by the corporate code monkeys and their shitty code ...

Hurd is not an option as for now, because f.e. it lacks a support for commonly used filesystems, and additionally I don't think that it will be able to become a serious alternative to Linux kernel without having an option to use proprietary firmware...

.

GNU Guix can be an option as it uses libre kernel which I believe does not include proprietary bits in it.