We imagine a future operating system

It would look like Windows XP with he security of Windows 10.

>>61829858
react OS 2019
>have faith

>>61829754
I want a terminal that can do simple shit without needing other softwares
like the one in templeOS, not even kidding, it's great

>>61829754
>But there is tech like SSD which at the moment doesn't seem to have enough specific software ready to cater for the change
Elaborate please. I have no idea what you are talking about.
How does currently available sofware limit currently and soon available hardware? You mean like incompatibility with high resolution displays?

>>61830004
Well specifically talking about SSD, shit like ext4 isn't built with SSD in mind. The only fs I can think of built with SSD in mind is F2FS. And the support for that is currently terrible in mainstream Linux.

High res displays could be another one though, DEs don't scale DPI well, you have something like Macshit which is customised to fit a certain display but out of that little walled garden, you get Linux DEs that look shit on new displays because the support is really fucking bad and things haven't been planned around new hardware. I'm hoping Wayland will change that for the better.

>>61829754
pure microkernel bsd

>>61830148
>shit like ext4 isn't built with SSD in mind
You're talking out of your ass. Not only are ext4 and xfs currently the only two file systems that support DAX, but NVMe is optimised for doing SG operations all over memory, which is makes it more than suitable for ext4.

>And the support for that is currently terrible in mainstream Linux.
That's also wrong. NVMe implementations in Linux are from Intel and are state of the art. For example, Linux was the first OS to get a NVMeoF implementation.

https://www.kernel.org/doc/Documentation/filesystems/dax.txt

t. someone who's actually implemented a NVMe driver

>>61830184
F2FS doesn't even fucking work atm, you can't install to a F2FS without major wanking around.

>>61830219
I wasn't talking about F2FS, I was talking about your bullshit claim that ext4 somehow works bad with NVMe SSDs, which is simply not true.

>>61830148
Okay, so File Systems optimized for SSD; GUI scaling for High DPI. These certainly are valid, but the improvement seems almost irrelevant.
What else is there? Network transparency and "everything is a file", i.e. Plan9. These are design improvements, sure. But what we have is (unfortunately) just good enough, and again not a very big step, I think. Then I could think of a less performance-costly virtualization technology, although there were a lot of improvements in that area in the last few years.
Honestly, I don't think there are very big issues where software holds back cutting edge hardware at all.

On the other hand, there are big political and market forces that impede a better designed, network transparent environment.
The trend in consumer electronics making everything into a service, and controlling it in a walled garden, together with an increasingly invasive surveillance from governments and increasing data mining from corporations is a huge issue, and I don't believe that it's going away any time soon.

>>61830241
I said it wasn't built with SSD in mind you tool.

>>61830265
That's irrelevant, because NVMe was designed with existing file systems in mind.

>>61830255
Also, now that I think about it, it seems like the opposite (current hardware limiting software) is quite prevalent:

Especially battery technology or rather energy efficiency in smartphone hardware (CPU, display) is nowhere where the market wants it to be. Same for VR applications. I can think of a lot more aspects where the hardware limits software design than vice versa.

>>61829754
I think in the computers general direction and it does what I want.