OSDEV, destroying all operating systems

I understand that developing a desktop operating system is a monumental task that even full time teams struggle with. Despite that, I hate all modern operating systems. I hate the proprietary and corporate nature of Windows. I hate the inconsistency and bloat of Linux. I hate the dichotomy between Qt and GTK with all the various "desktop environments". I hate the lack of simplicity and control over my system, even on "free distros" you're only hacking within a sandbox because you're forced to use the software like the developers intended and forking the software is usually not worth the effort. Due to the open nature of FOSS, when multiple developers start contributing each one has their own vision, tries to take the software in their own direction with little understanding in regards to other contributors and their code. This is technological chaos which leads to bugs, security vulnerabilities, inconsistencies and incompatibility. I enjoy OpenBSD but it's obviously not designed or ready for the desktop and if it was it might slowly turn into another Linux. The OS would be built as close to the bare metal as possible, with a minimal skeleton codebase that can readily and easily be recompiled and built on top of, including the skeleton itself if you know what you're doing. I want to reinvent the wheel. It can always be rounder. Anyone who disagrees has corporate brainrot and is domesticated.

is it going to be a unix like? and what language are you going to use?

>>66405 I'm not sure yet. There can be a translation layer if it isn't. Likely C. If you don't like C you can make wrappers for your preferred language.

>>66404 Why don't you contribute to Haiku? It seems pretty cool project! >https://www.haiku-os.org/about/ >>66405 >is it going to be a unix like? I really hope it isn't because there are a lot of unix-like operating systems and I think it would be much more fun to explore alternative designs. If you want some food for thought, please watch: https://vid.puffyan.us/watch?v=7RNbIEJvjUA (#rC3 - What have we lost?). I think TempleOS is unironically more interesting than most OSDev projects because of this. But I recommend you don't invent your own programming language, if you want contributors to your project. >Site: https://web.archive.org/web/20170221041504/http://www.templeos.org/ >iso archive: https://archive.org/details/TempleOS_ISO_Archive (some versions lack After Egypt) >docs: https://templeos.holyc.xyz/Wb/Doc/HelpIndex.html >vids: https://archive.org/download/TerryADavis_TempleOS_Archive/videos/ (and other archives) (There are also forks like Shrine and ZenithOS) How to install? #1. Install qemu and download the TOS iso. #2. Pick a qemu script that Terry wrote and adapt it to your needs: https://templeos.holyc.xyz/Wb/Home/Sup1/Sup1Bin/QEMU/ # Example: # Create a virtual HDD: qemu-img create -f qcow2 TOS_disk.qcow2 3G # Install TOS qemu-system-x86_64 -hda TOS_disk.qcow2 -smp cores=2 -enable-kvm -cpu host -m 2048M -rtc base=localtime -soundhw pcspk -cdrom TempleOSCDV5.03.ISO -boot d # Run the VM qemu-system-x86_64 -hda TOS_disk.qcow2 -smp cores=2 -enable-kvm -cpu host -m 4096M -rtc base=localtime -soundhw pcspk -boot c

>>66415 I've heard about Haiku but don't know much about it. I'm not only trying to develop a new operating system, but a new philosophy aswell. That's why the subject is titled "destroying all operating systems". Because from the point of view of this new OS, all existing systems and protocols will be essentially deprecated, only kept around as legacy support.

>>66405 >is it going to be a unix like? Please, no. There are a fuarrrkton of unix-like out there and people will just keep using Linunx anyway. I want more experimental cool soykaf.

>>66422 You should look into TempleOS, and I'm not even kidding at all - TempleOS actually has a lot of brilliant ideas and design decisions for your use case, personal computing, like having the shell use the premier programming language of the system https://www.youtube.com/watch?v=d3eFHyryopQ If you actually want to do this, I suggest that you aggressively limit your scope, and try to recycle components wherever possible - not only does this mean you don't have to do as much, it can actually lead to a more cohesive environment. Whatever you end up doing, best of luck.

>>66404 Do you have any actual plans for developing an OS and its microcomponents (bootloader, kernel, init, libc, drivers, coreutils, gui server, widgets, etc)? I can offer some hints (wrote an OS myself) but you need to do a roadmap for it first. What's the first step?

>>66404 Openbsd is ready, you arent. Thats the real truth.

Your probably isn't operating systems, it's neuroticism. Limiting technology to what one person can rebuild, results in something "not ready for the desktop". It's inherently too complicated a task for one person.

>>66431 I thought I'd start off by doing some experiments on my existing OS, by making prototype environments. >>66432 Adults are talking. >>66434 I said "readily and easily recompilable". That means it builds itself.

To elaborate and clarify, I hate most technology both old and new. As Steve Jobs puts it; "We're creating tools for the mind." Computers have become an extension of the mind as it manipulates our perception of the world, computers have become a form of control. We must have full control over every aspect of our computers, at least to the best of our ability. But that isn't possible with existing technology because of the errors inherit to modernity that inhibits our creative abilities. The purpose of this new operating system is to break every convention or norm and put thought to practice. It's a breaking point, a chance to reset and start over. One idea for example I really want to try is a mesh network over radio for an internet free, truly decentralized communication system.

>>66443 >a mesh network over radio That already exists. You also can't encrypt radio communication without getting in trouble with the feds. This is what happens when you don't learn from others, you retread the same ground.

>>66435 >adults are talking No. If you were an adult you wouldn't have an issue with OpenBSD. Its the best operating system and you have to be ready to compromise to use it. You choose hardware according to the OS, you dont expect an OS to cater to your hardware. This is what an adult OS is.

>>66450 Except no hardware targets memebsd. You mean old hardware. https://isopenbsdsecu.re/quotes/

>>66451 The idea that hardware should "target" an operating system is deeply evil and perpetrated by Microsoft and Apple.

>>66452 Who is going to buy hardware no OS has drivers for? Or should the hardware manufacturers writer drivers for every memeos?

I'm interested to hear more about your goals for this project. >>66443 What do you see as the problems with modernity? What do you want to achieve through a free decentralized communication system that hasn't been done in other ways? Many projects with this as a goal end up populated solely with tech enthusiasts interested in the tech itself that were already talking on other platforms or sites, rarely does this reach a more general audience. Do you have ideas on how to address this or is this not a goal of yours at all?

>>66456 I'm interested in that too but after seeing so much larping on this subject, I don't think we're going to see a single line of code.

>>66450 I agree OpenBSD is the best we have, but it's not good enough for me. I do intend to use a BSD license and take some inspiration. >>66452 Software should be built to suit the qualifications of the hardware. >>66456 Because of how the internet is designed, it's not possible to have a truly decentralized network (without expecting every user to know how to port forward). The two major problems that lead to that specific idea are centralization and complexity. Unfortunately, to gain a user base outside the tech nerds sphere specifically if it's to rival any big operating system, it might need at least semi corporate backing. I may set such a branch on a course for its own self destruction once it is no longer needed.

Someone posted this on 4chan in a thread about OSDEV. http://doc.cat-v.org/bell_labs/utah2000/utah2000.pdf I'm currently working on a paper detailing my goals for this project as currently it's a vague "anything but this". Once I am confident with a more specific vision, development can begin. In essence we're going back to 1970, starting again from scratch. Like a mathematical equation it must be perfect. If you get the answer wrong for X, it doesn't matter how smart you are. The mistake will cause the result to stray further from the true answer with every passive iteration. It's demotivating knowing this project will likely become something obscure that no one knows about. I will still try despite any doubts. If this actually does grow into something, we can become masters of our software. It can become the ultimate power tool for hackers and laymen alike.

>>66526 I saw that paper posted there too, it was pretty interesting. Check out the talk posted there as well. Hope to see your paper soon.

>>66526 >rob pike not reading that

>>66404 I like your enthusiasm, please do not give up

>>66532 Is there something I should know about him? >>66533 Thank you!