-
Notifications
You must be signed in to change notification settings - Fork 1
Home
CrystalOS Wiki
the initial aim of this project was to follow a blog series on how to make a custom operating system found here:
with the github repo for his project here:
https://github.com/phil-opp/blog_os
After reading and implementing the features from the final chapter, (async/await) I could find no further instruction on how to continue with the project from there despite the author of the series saying over a year previously that there would be more posts coming soon.
while I'm waiting for the third edition to release, I guess I'm gonna just have to improvise :)
the blog got me through the memory management side of the process so i believe that I should have a lot more breathing room to implement the features that i want. As of completing the tutorial, i obviously still dont have access to a standard library, however i can at least use Vectors and Strings now which are important types, as well as the fact that i have access to async and heap allocation.
- whenever i have the chance to work on this project, i want to try and implement a new utility which could be useful or cool for anyone using the operating system.
- this could be anything from a cool neofetch style ascii fetcher (if you dont know what im talking about, its just a cool ascii logo of the operating system that appears when you open a terminal sometimes)
- improve the text rendering system to create a set of globally accessible functions and/or macros in order to render the text in a more visually appealing way to the user (as the default yellow text does look extremely ugly lmao)
- implement a basic text editor (this will be difficult)
- i would need a way to move the cursor around the screen and print text at that location
- this would mean rewriting the majority of the code for the vga buffer module to create a more flexible system which allows for applications (modules / commands) to take more direct control of the text rendering whenever they are active
diverging from the original blog series, i have made some significant changes to keyboard.rs
- i have moved the source code that handles the keyboard input from keyboard.rs to shell.rs
- this means that instead of the operating system running a task on startup that continually awaits a the next keystroke and works from there, the new layout works very differently
- firstly, i use a lazy_static creating a static called CMD which houses the shell itself
- this allows me to reference it from anywhere in the code and initialise it as soon as the program runs
- this may be changed later as i could just make an init function in shell.rs if i needed to
- the shell contains a get_input function that awaits a keystroke from the user before continuing
- this is looped inside the main shell function and added to a buffer
- when the \n character is inputted, the buffer is copied to the command history vector and then cleared
- additionally the buffer is run through a match statement that will start any app that matches the command or alias.
the crystal api will essentially be a standard library for any programs that are run by the shell
- it provides basic functions such as waiting for a keystroke or string to be entered by the user
- it will eventually support coloured text output once ive had a chance to modify the code for the vga buffer to support coloured text output through a public function.
here is a template that could be used to program using the crystal API
// ignore everything from this point up until the App struct
// --------------OS-INTERFACE-------------------------------------------------------------------------------------------------------
use std::io;
use std::io::Write; // ignore these, i have my own implementations that i will replace them with
struct CommandHandler {} // a struct used in my code (just ignore)
impl CommandHandler { // dont modify anything here
fn new() -> Self {
Self {}
}
fn input(&mut self) -> String { // this function will get replaced by the custom input function
let mut string = String::new();
io::stdin().read_line(&mut string).expect("error getting input");
string
}
}
fn main() { // the entry point to your code, it calls the code for the application
// will be removed when integrated into the os and replaced by the shell command
println!("");
print!("enter arguments to run command with > ");
io::stdout().flush();
let mut args = String::new();
io::stdin().read_line(&mut args).expect("failed to get input");
let mut app = App::new(CommandHandler::new());
app.run(args);
}
// --------------IMPLEMENTATION-----------------------------------------------------------------------------------------------------
struct App { // change name to whatever you want
handler: CommandHandler,
// any global variables for the application should be put here
// in the form: varname: VarType,
}
impl App { // name must be the same as the name of the struct
fn new(handler: CommandHandler) -> Self {
Self { // this should add any variables that are needed while the application is running
handler: handler,
// status: String, (example)
}
}
fn input(&mut self) -> String { // this function gives command line input
self.handler.input()
}
fn run(&mut self, args: String) -> Result<(), String> { /*
this represents your actual main function
write all the code for your program starting here
use println!() to print to the screen
use self.input() to get input from terminal
*/
println!("app running {}", args); // do stuff here
// example of how you can use the input function
println!("type something");
println!("input: {}", self.input());
// if you want to return an error, write: return Err("error message")
// the error message tells the operating system what went wrong with the code or user input.
// if you want to return ok, write: return Ok(()) (make sure to have the 2 sets of brackets)
Ok(())
}
}eventually i want to try rewriting the majority of the code for the VGA buffer. this is so that i can implement what i'll call a 'sandbox mode' for the screen. this mode will support:
- moving the cursor around with arrow keys
- writing text at the cursor
- writing coloured text anywhere
- reading the entire output of the vga buffer or just a line into a string eventually, this could theoretically lead to a library that was able to support things like a basic text editor for writing out messages and the capability to theoretically program basic 2d games in an ascii art style (something like space invaders, tetris, etc.)
- created a standard library of functions that any application can use
- implemented a random function to the standard library that can generate random numbers in a range
- added the print and println macros to be directly accessible from the standard library
- added global functions for getting input as a keystroke or from the command line
pub async fn stdin() -> String // returns the string inputted by the user
pub async fn stdchar() -> char // returns the next keystroke
pub fn crate::std::random::Random::int(min: usize, max: usize) -> usize // returns random integer in range (min <= x <= max)
pub fn crate::std::random::Random::selection<T: Clone>(list: Vec<T>) -> T // returns random element of vector argumentsince the last update i did a few things.
-
refactored the entire codebase, moving the standard library, kernel and applications to separate folders in the source code in order to better organise them
- this should make my next goal much more seamless - this goal will be to further abstract the applications from the kernel, essentially there will be a lib crate containing the kernel and standard library in addition to a binary crate which actually has the bootloader / init system / applications etc
-
wrote a basic ASCII rendering engine that can create and place 2d element anywhere on the screen
- the elements are placed using a coordinate system where the top left character of the element is placed at that coordinate on the screen
pub struct crate::std::io::Elementafter taking a bit of a break to get my A level work done and play some games, i've finally found time to get back into this project
over the last couple of days I've made some signifcant changes to this project. I have started the development of a TUI (terminal user interface) library allowing for widgets to be displayed on screen using ascii. its pretty barebones at the moment but I'm hoping to add more features as time goes on!