Can you still do what he did or is developing something like the Linux kernel from scratch just not possible anymore in...

Can you still do what he did or is developing something like the Linux kernel from scratch just not possible anymore in this day and age?

Attached: torvalds.jpg (800x1113, 128.72K)

The technology isn't there anymore :^(

Explain

The knowledge of the dark arts required to manifest a kernel into creation have been lost to time; we can only build upon the foundations our forefathers laid.

Complexity has grown at the kernel layer to the point that it is just barely manageable, and complexity has since creeped up higher in the stack as it grew. You can build a basic real mode kernel nowadays, but the chances of it being useful in any way are extremely low. Literally hundreds of thousands / millions of engineering hours have gone just into optimizations for Linux, let alone features.

it was never possible. the linux kernel, as we know it today, only exists because of the contributions of tons of big companies and people over the course of 20+ years. it's not something that torvalds did singlehandely or overnight.
what torvalds did was create a (relatively) simple kernel and publish it with a freedom license.
stallman wanted a 100% free operating system, but didn't have a kernel, so he used linux.
more and more people started contributing to linux, and the project grew.
since no one likes reinventing the wheel, and the kernel was free to use, people started using linux in places where they needed a kernel (a good recent example of this is google's android. google wanted to make a phone os but rather than building their own kernel they just used linux since it was good enough for the task). this in turn led to more popularity and more contributions to the kernel.
torvalds' success is mostly about him being in the right place at the right time. writing a kernel is certainly nothing to sneeze at, but the kernel torvalds wrote is hardly the kernel linux is today. it only got to the point it is at today because he got lucky and it got adopted by many big companies.
if you define "what he did" as write a kernel, then yes of course you could still do it today if you wanted to.
if you define "what he did" as start a project that would eventually be the kernel of choice for 100% of the top 500 supercomputers, 98% of the top 1million websites, and every android phone then no. that ship has sailed.

There's at least 50 different usable OS kernels in active development but not much in the way of userspace for any of them. #1 thing people want is a capable web browser so you're effectively limited to platforms they support, or platforms that can emulate an environment they support.

So do you think that we will always use the Linux kernel? Will another kernel EVER be developed again?

No, the incantations required to turn regular code into kernel code are lost. Once Jim Keller dies, the knowledge of the dark arts for CPU design will be lost too.

Attached: 1630217128053.png (1440x2141, 2.07M)

I'm still rooting for GNU Hurd