The logo is already done btw
the idea is:
Bootloader in assembly x86
the kernel is in C
a mono-kernel
implementation of unix system
implementation of the windows kernel and the osx without bloat
no windows, only tilling as a gui
fuck bash, we will use fish
lets go!!
Other urls found in this thread:
youtube.com
codersnotes.com
twitter.com
This is the start:
i stole it from github, its mine now
should we remove the UI? i3 is better
First, we need a way so we can collaborate, we're going to use GitHub, I guess? Who's hosting? Also, what's it going to be c Mockups?
Listen you dumb, have you ever made your own compiler? I made one, in Assembly x86 5 years ago, do you even know how to code?
it's boring to use github, just give me the code here i will paste it somewhere in the code
At the very least post the source code somewhere?
what do you mean? It's not open-source, everything is mine, just give me the code and say me what to do, I will paste it using CTRL-V
I'm ok with the use of "goto" that's not a problem!
> The logo is already done
Dropped.
>It's not open-source
Aaaaaaand closing tab
We're not going to give this shit to you for free for you to profit from
>implementation of the windows kernel
Go grab the ReactOS code, then proceed to kys
That looks pretty good actually, what theme is that?
OH BOY, IS THIS THE NEW TOX? OH BOY, TIME FOR 500 THREADS OF INFIGHTING ABOUT MARKETING
>It's not open-source
That's where this whole thing failed as a "community project", atleast for me.
You can't have others develop software unless they have full access to the same source code as you. If you want to make something that's "not open source", you should still be using git, just with a private repository. That said, you will also need to give people a reason to contribute if you want to make the source code all to your own. Will you be paying us?
dont use package managers
package managers are for cucks who dont run .exe's off of media fire
i remember this user reporting on his OS in every DPT, and then it blew up on hackernews and now he's too good for us
>unix implementation of windows kernel
Imagine being this retarded
>make me a os so i can profit form your work
>The logo is already done btw
Well that's half of the job done then
Serenity user is still in /dpt/ every now and then but he got a job and stopped NEETing
>The logo is already done btw
Typical Linux priorities.
Now leave the GUI to the last.
why does world need another carbon-copy of unix?
>x86
lol
ARM or get out
If you've never wrote an OS before, then start with UEFI unless you want to fucking suffer
Here's the first program for your OS kernel, OP
// Author : user
// Date : 6/7/19
// Filename : kernelcode.c
#include
int main() {
printf("Let it be known that OP is a faggot.");
return 0;
}
Please add it to the kernel OP! :3 UwU
Unironically TempleOS's features are so fucking revolutionary it's mind blowing.
Unix is for teletype terminals and mainframes.
Think actual personal computers.
We should be mashing up the useful stuff from Temple and Plan9
EFI version:
// Author: me
// Date: 6/7/19
// Filename: kernelcodeefi.c
#include
#include
int main(void) {
ST->ConOut->OutputString(0x0F, L"Let it be known, that OP is a massive faggot.");
return 0;
}
>Sat
I AM THE SMARTEST PROGRAMMER THAT EVER LIVED
the logo isn't minimal enough
>Unironically TempleOS's features are so fucking revolutionary it's mind blowing.
such as? I acknowledge that unix is trash is nearly every way, but I failed to notice anything *revolutionary* on TempleOS.
Do it OP
It'll be your own proprietary code
I’m still here user. :)
>that owl
>17KB
>minimal
choose one
I'll bite the bait:
* gui, bash/fish - you are thinking too high for now
* windows kernel - good luck, Reactos is in making for a looong time and not anywhere near finished product. Also * osx - ok, kernel itself is opensource, userspace is not
* impl. of Unix or something similar - sure
* monolithic kernel - wouldn't micro or hybrid be better/easier to do without the budget Linux kernel has?
* kernel will probably also have some assembly inside, I'd suggest C or Rust for most of the things, although on that low level Rust safety claims would be severly limited (bo standard library and all that jazz)
* I wouldn't worry about bootloading from BIOS/UEFI right now, let's say for first version everything woukd be ok as long as GRUB (or SYSLINUX) could boot the kernel.
* make some parts of the kernel platform agnostic so it could be expanded to other archs than x86
As for the fin stuff: Logo/name should be assigned in competetive matter: everyone can suggest logo and name, but the person who first implementa something (e.g. keyboard support and vga text output) chooses respectively name and logo from all suggested ones
As for the codename FPBS, so it would be "qsdqd"
Suggested first steps:
* create public qsdqd repo
* make enough code to push "hello world" through serial on boot
* planning: decide how exactly should the kernel look like inside (important step, you can fuck up everything there)
* work from that adhering to the plan from previous step
So pretty much the same old shit, nothing innovative. Nah, fuck off.
What do you mean?
>The logo is already done btw
i have no interest in this project
It's fuucking stock image, cannot be used unless you want to get cease and desist or other shit like that
After short thinking:
* maybe make first version boot from PXE, that way we could compile it on one PC and easily run it in VM AND real harware, no copying between machines
* Planning should also contain choosing environment tools (gcc9 plus Intel or AT&T assembly), documenting everything and creating scripts/Makefiles/Cmakelists.txt for compilation
Make a repo then
Why not have the kernal in rust?
Andreas still posts you fucking cuck. He doesn't whore attention like other retard devs. He's a nice guy when you hang out on IRC
Read this, it's relatively short but makes it clear, esp in second half of the page
I know what TempleOS, I've been using playing with it for some time, I've read tons of its source code. I've also read this article.
I get that unix is a stockholm syndrome OS and unix standards were literally created to freeze any innovation in 90s. But really, there isn't much of an innovation there. It's riddles with lisp and forth weenies concepts and designs on top of C-like language, so this decision is kinda unique.
It's the [birthrates] DolDoc, dude
I dare you to find me an "interactive" shell on *nix or anywhere else that can handle images, hyperlinked wiki-like executable rich text as well as flowchart-in-source.
The only thing as wild and liberating as that is the way plan9 does everything-is-a-sharable-networked-file-including-hardware (and actually DOES IT unlike any other *nix)
emacs, web browsers
So, you're saying 100 layers of crufty api's and web-dependent systems cobbled up from similar piles of layered bullshit using 2000% more hardware resources is equivalent to being able to do it everywhere in the OS, offline and natively and there's nothing to see here?
everything is cleaner when done from scratch
xanadu project is pretty revolutionary, TempleOS is just less convoluted than what you normally see
This thread has just the right amount of retarded larping
Austism
We're not giving you code for free, pay us then maybe.
>a mono-kernel
already off to a horrible start