Linux 4.17 Change To Allow RTCs To Live Beyond Their Intended Life

[mulenmar]

I'm pretty sure consumer IoT won't last past 20 years (or less) due to planned obsolescence and cheap parts.

I've seen so much crap not surviving 3 years...

I doubt it will last either, but I see no reason it won't be getting newly-manufactured and then installed on the orders of Pointy-Haired Bosses in 80 years.

[starshipeleven]

My guess: 100 Earth years from now, the only things still running Linux will be old IoT devices nobody bothered to upgrade, linked into botnets slowly mining Megabytecoin.

I'm pretty sure consumer IoT won't last past 20 years (or less) due to planned obsolescence and cheap parts.

I've seen so much crap not surviving 3 years...

[mulenmar]

My guess: 100 Earth years from now, the only things still running Linux will be old IoT devices nobody bothered to upgrade, linked into botnets slowly mining Megabytecoin.

My hope: typical systems will be running seL4 kernel-based OSes with userspaces written in Rust.

My fear: typical systems will be running dataMinerOS 13.37 NSA-Interpol-neoKGB Edition, impossible to install anything else on due to UEFI non-disableable Secure Boot whatever-point-0. Assuming we're not all using slightly-radioactive abacuses by that point, of course. Both seem equally unlikely.

[starshipeleven]

You don't hear a lot of people talking about it but IBM mainframes still run a fair chunk of the ol' internet.

Also, AS/400 (or their successors, still called "AS/400" by the people working with them even if they are IBM i Systems or IBM Power) are still working as database servers in many places I know.

When you see a relatively complex ncurses-like interfaces on a sysadmin console, that's one of them on the other end of the wire.

[bridgman]

Since we aren't using any of the OS's from even 30 years ago, I'd have to guess no.

As far as I know IBM's z/OS is pretty much a direct evolution of the mainframe OS code we were running in the 60's - still supporting all the same application runtimes, and still running old applications. The main change is supporting wider virtual addresses, going from 24-bit to 31-bit and now 64-bit.

You don't hear a lot of people talking about it but IBM mainframes still run a fair chunk of the ol' internet.

A bigger question is whether the US will still be flying B-52 bombers a hundred years from now, smoky take-offs and all.

[starshipeleven]

Since we aren't using any of the OS's from even 30 years ago, I'd have to guess no.

That's a bit simplistic for a long-term speculation. Back then the market was new, it's normal that many of the competitors died off when there were many and the competition was fierce. There were the Unix Wars for example.

Linux itself is 27 years old already, FreeBSD is like 25 or so. Windows (NT kernel) is 25 years old, MacOS is 18-ish years old.

There is nothing seriously threatening Linux and FreeBSD, Windows is on a long slide to hell but it will still take a very long time to die off and there is still a big window of opportunity for rejuvenating it if MS finds their brains, meanwhile people buying Apple devices to do the same 4-5 things while looking trendy will still exist in the future.

The only thing that can change the current situation is some breakthrough that makes the current OSes incompatible with next gen computing hardware, or something equally big.

Can it happen? sure, over a century anything can happen all-right. But that's not something you can count on.

[starshipeleven]

I wonder if Linux will still be around as an actively used kernel 100 years from now.

Assuming that there are no dramatic changes in computing technology, then Linux is the kernel that has the most chances to survive till then, due to its adaptability, because it is not locked to a single vendor or CPU architecture, and because it has cornered the market already in many different segments and it would be a PITA to dislodge it from there.

Windows, MacOS, and most other OS/kernels are much more limited in scope, and are tied to a single vendor. They have their own strongholds too but they have little chance to spread to more fields more than they already cover by now. Linux can lose a market, and many vendors, and it will still be fine and will still have the ability to get back, most other kernels are bound to a specific OS which is again bound to a specific vendor and specific market.

Although in 100 years you can very well have some dramatic event, technology breakthrough, or whatever, and in that case all bets are off.

[George99]

Since we aren't using any of the OS's from even 30 years ago, I'd have to guess no.

From time to time I boot into Amiga Workbench from 1988. I don't own any Commodore Amiga any more, just using the FS-UAE emulator.

[Zan Lynx]

Since we aren't using any of the OS's from even 30 years ago, I'd have to guess no.

https://en.wikipedia.org/wiki/FreeBSD
The BSD project was founded in 1976 by Bill Joy.

And BSD UNIX is a direct ancestor of Mac OS X Darwin.

I do think we'll be using something called Linux in 100 years. How much it will resemble the Linux of 2018 is unknown. It might have been modified so much that it's unrecognizable. Heck, maybe Linux 2118 uses machine learning and rewrites itself to match the provided hardware as well as any other hardware it detects and has authorization to use inside the local quantum mesh.

[NateHubbard]

I wonder if Linux will still be around as an actively used kernel 100 years from now.

Since we aren't using any of the OS's from even 30 years ago, I'd have to guess no.