diff options
Diffstat (limited to 'comp/gift/writeup.tex')
| -rw-r--r-- | comp/gift/writeup.tex | 339 |
1 files changed, 339 insertions, 0 deletions
diff --git a/comp/gift/writeup.tex b/comp/gift/writeup.tex new file mode 100644 index 0000000..74bf0d8 --- /dev/null +++ b/comp/gift/writeup.tex @@ -0,0 +1,339 @@ +\documentclass[a4paper,12pt]{article} + +\usepackage{geometry} +\usepackage{titling} +\usepackage{titlesec} +\usepackage[english]{babel} +\usepackage[hidelinks]{hyperref} +\usepackage{listings} +\usepackage{xcolor} +\usepackage{graphicx} +\usepackage[export]{adjustbox} +\usepackage{forest} +\usepackage{tikz-qtree} +\usepackage{bchart} +\usepackage[T1]{fontenc} + +\titleformat{\section}{\Huge}{}{0em}{}[\titlerule] +\geometry{a4paper, total={170mm,257mm}, left=30mm, right=30mm} + +\author{Lucas Standen} +\title{A parting gift of my knowledge} + +\begin{document} + +\maketitle +\newpage +\tableofcontents + +\setlength{\parskip}{1em} +{\setlength{\parindent}{0cm} + +\section{Headnote} +I leave this as a gift and thanks for all the things I learnt in your +class, hopefully you can learn something from this. This document +contains almost all the things I know about computer science; please +enjoy! Not much of this will be stuff needed to teach the course but I +still thought it would be nice to give you a wealth knowledge for the +fun of it; If knowing me for 2 years wasn't enough tell, doing things +for the fun of it is the best way to learn (in my opinion). + +I will be laying this document out into sections on general +topics and then subsections of specifics, see the table of +contents to do further reading. + +\section{The Unix wizard handbook} +\subsection{The history} +Unix was originally made at AT and T's bell labs, their research centre +at the time (It belongs to Nokia now). It was designed as a small and +lightweight OS for the 'weaker' computer at the site although it still +took up and entire room. It was made by Ken Thompson and Denis Richie in +the early 70's. They built everything from the ground up for the system, +it was written in C, which they made specifically to write the system. + +It is often cited, but we will never truly know, that Ken was +the person with all the ideas, and Denis was very good at +implementing them, so most of the code in the original Unix was +written by Denis. I don't know if there is anything relevant in +that, but I thought it was interesting. + +Unix was sold to many university's in the late 70's and early +80's, and it was through this that it really started to gain +some traction with the students. Many developed Unix further, +because unlike modern paid software, Unix was distributed in +source form, meaning you could tinker it as much as you liked, +assuming you had brought a license. One of the main outcome of +this would be BSD (The Bearkly software distribution) version of +Unix. The people who worked on this system later went on to +develop Java (All of computer science is so interlinked, its a +small world!). + +In the 80/90's Unix was de facto in Uni's but the free software +movement was starting to gain traction, and many re-writes or +clones of Unix were starting to show up, like GNU (fun fact, GNU +is a recursive acronym for GNU's not Unix!) + +\subsection{The philosophy} +The main idea's behind Unix were that it needed to be small to run on +the 'weaker' computer at bell labs. One of the leading limitations with +this system was its small storage size of about 20Mb. To avoid this +being an issue they came up with the following philosophy to ensure all +software remained small, and simple even on the weaker computer. + +\begin{description} \item[Small reusable tools] The idea was + that if + you made one small program that did one thing and provided the + ability to combine programs easily, you could let the user put + together it together to make the tools they needed to their own + needs. + \item[Everything is a file] This was the idea that if everything + is a file, from talking between systems, to talking to kernel, + you could avoid many different APIs that would repetitive, + instead all you needed was one single file and you could work + with the entire system. +\end{description} + +Something that was more unintentional but still worth noting is the Unix +was designed for Teletypes, (paper printers that talked to computers), +that didn't have the ability to overwrite sections of the page, and +printed at a slow 10 characters per second. This effects a large amount +of the user interfaces for Unix. + +\subsubsection{GNU/Linux} +One of the many outcomes of the free software movement, it quickly +became the easiest way for anyone to use Unix, releasing in 1991. There +were open and free versions available at the time, such as BSD, however +you still needed to be in a university to get a copy, as the internet +wasn't really active at this time. It is ironic that Linux became a core +part of this system, as it was in fact a small hobby project that the +creator Linus Torvalds said 'this is just a small thing, it won't go +anywhere'. + +The GNU part of GNU/Linux is the user land utility's of the +system. Linux is only the kernel of the system, handling the low +level system, such as the file system, memory allocation and +process scheduler, the rest of the GNU/Linux system is the GNU +core utils, such as the GNU bash shell. + +\subsection{The shell} +\subsubsection{The interactive shell} +The shell is the part of the system that you time commands into, it's a +whole programming language that can run line by line (if +it helps you think about it, its like using the python +shell in IDLE). + +The shell was designed to be a helpful interface to take the +user in and out of programs, it is fast and simple. So here is a +whistle stop tour of all the things it does. + +The shell can store variables that contain strings, these +strings can then be accessed by any program that is run in the +same shell that the variable is in (think of it like a +parent-child thing, the child process knows all the same things +as the parent shell). The can be created like so 'myvar="hello"' +and then can be accessed using a '\$' like so '\$myvar', this +syntax is also seen in Perl and php, which are often compared to +the Unix shell as they are all interpreted scripting languages. + +On a Unix system applications are stored in /bin, /usr/bin, +/usr/local/bin, and \url{~/.local/bin} (and sometime other +places too). The shell will allow you to type a commands name, +like 'cp' (copy paste), and it will search through all the known +locations of programs and run the first one it finds. The shell +stores all the known locations of programs in a variable called +\$PATH. + +Programs can be started in parallel using a '\&' sign, which can +be very helpful when testing networked software. + +Modern shells such as Bash and Zsh provide autocompletion of +possible commands when tab is pressed, just like modern code +IDE's allowing you to auto complete code. + +Commands are always processed in the following structure +COMMAND-NAME -ARG -ARG -ARG [FILE] [...] Some commands may take +a fixed amount of arguments, while others maybe more variable, +some may take in none. Most commands will perform some +operations on a file. + +Almost every command can be run like so COMMAND-NAME -h or +COMMAND-NAME --help to get a list of available options and how +to use it. + +Arguments that are only 1 letter long will usually only use a +single - and arguments that are words will usually use a double +--. + +The shell has a current working directory, this is a folder that +it is currently navigated too. When typing commands, file +arguments can be given as a relative path to the current working +directory. For example 'cp file1 file2' will copy file1 to +file2, where both files are in the same directory. You can also +use absolute paths which always start with '/'. + +There are 3 magic shell paths, ../ which is equivalent to the +parent folder of the current working directory, ./ which is the +current working directory and \url{~/} which is the currently +logged in users home director (where their documents, downloads, +videos, etc folders are stored). + +\subsubsection{The common commands} +Items given in brackets are optional arguments +\begin{description} + \item{cp file1 file2} \\ cp will copy file 1 to the file 2, if file 2 + already exists it is overwritten, if it doesn't exist created. The -r + option is often used to allow it to co instead of individual files. + + \item{rm file1 [fileN]} \\ rm will delete any files provided, it + is often used with the -r flag like cp is removed folders as + well as files. The -f option is also used to avoid prompts of + weather you want to delete the just deleting it straight away + + \item{mv file1 [fileN] file2} \\ mv moves 1 file to another + location, it is equivalent to copying a file then deleting + original. For some reason unlike rm and cp, it doesn't the -r + option to work with folders. If file2 is afolder (something + that ends in /) multiple files can be given and they will all + copy to the folder. + + \item{cd [folder]} \\ cd will set the shell's current working + directory to the folder specified. You can use this wi magic + paths, for example cd .. will move you back one directory, and + cd \url{~} will move you to you home directory. Running cd + without any arguments is the same as cd \url{~} + + \item{ls [folder]} \\ ls will show the folders and files that are + contained with in the folder provided. If no folder provide it will + default to the current working directory can use the -l option to get + file permissions and the -a show hidden files. (you can think of cd and + ls equivalent clicking on a folder in file explorer and observing th + contents). + + \item{cat file1 [fileN]} \\ Cat is a simple command that will + print the contents o files given to it as arguments. + + \item{grep regex [file]} \\ Grep is one of the larger commands + and has + many abilities however its main one is to search through a file + for a + expression. If no file is given to the command, it will read + from standard input (typing input) + \item {sed expression [file]} \\ sed is a find and replace of a + file (or stdin if no file is given), the expression should look + something like this. "s/word/newword/". the 's' at the start + tells you that it is in subsitution mode (other modes exist), + the word is the expression to search for, and the newword is wat + to replace it with. By default this will only substiute 1 word + per line, if you want to do it more times the g (global) flag + can be given like so, "s/word/newword/g". No matter how the file + is give, via stdin or a file name, the output will be printed to + stdout. +\end{description} + +\subsubsection{Pipes and redirections} +This is where the real magic Unix starts, as previously mentioned, Unix +was designed with the id of making a small reusable tools, and this is +how the tools can work together. + +One can take the output of one command pass it to another, or +copy the outputs to a file. (A commands input is what is typed +in by the user, and the commands output is what is printed to +the screen) This can be done like so. ls > files.txt This will +take the output of the 'ls' command and copy it to the file +'files.txt' (which is in the current working directory, we can +tell this because it doesn't have any '/' in its name). You can +also do: + +ls >> files.txt + +Which will do the same thing, however if the file already exists +it will append to it instead of truncating it. + +This on its own is quite helpful, as many commands take files as +arguments, and this will allow to use one command to help +another (the '>' and '>>' is called a redirection in Unix). + +One can also use a '|' to take the output of one command and use +it as the input to another. + +Here is an example: + +cat file.txt | grep [0-9]+ + +This has taken the output of the cat command, and put put it +into the input of the grep command (as if you had typed what the +cat command printed), and has append a special EOF character to +the end of the cat command output, to specify the end of file. + +Grep has then read this input, and performs an operation on it, +based on its arguments. + +\subsubsection{Shell scripting} +This is where you take all the things you've used in the shell, and wr +ite them into a file, as simple as it gets really. The only notable +differences is the fact that you have to prefix a file with a special +string to say it is a shell script and make it executable. + +This can be done like so: +\begin{lstlisting}[language=bash, caption=example.sh] + #!/bin/sh + sed "s/word/newword/g" myfile > tmp + myvar=$(grep "[0-9]+" tmp) + echo $myvar +\end{lstlisting} + +This simple example will run sed, and put its output in tmp, then run +grep on the output and then echo the output of grep. The \$() syntax is +used t o specify run the following value as a command and put its output +here. A mixture of sed and grep can create some rather insane programs, +in 2023 i use these two tools to do almost all of advent of code, by +taking the inpu t and then manipulating into valid python code, which +then solved the task for me . Sadly i dont have the scripts I used +anymore, but there are some great videos of Russ Cox (co creator of the +GO programming language) doing the same thing +\url{https://www.youtube.com/watch?v=hmq6veCFo0Y}. + +\subsection{Device files} +As mentioned prior, everything in unix is a file, this went all the wa y +down to devices and hardware attached to the system. Here is a quick +list of a ll the major files that you can use inside of programs to +communicate with th e hardware on a device. +\begin{description} + \item{/dev/random (/dev/urandom)} \\ These are both random text + generators that can be found on most unix systems. You can use commands + like head (another small unix tool that prints the first 10 lines of a + file), to read it like so + \begin{lstlisting}[language=bash, caption=/dev/random example] + key=\$(head /dev/random) + \end{lstlisting} + urandom works the same, although it uses a slower, more random + algorithum to generate text. + \item{/dev/null} \\ this is just a bottemless pit file, it + contains nothing, it is often used to compare if something is + empty, saying if the contents of a file is equal to /dev/null + then its empty. It can also be used to delete files if you + wanted, via mv file /dev/null (although rm does the same + thing). + + \item{/dev/sdXN} \\ this is how devices are listed, your + harddrive (or whatever the kernel detected first) is equal to + /dev/sda and then each partion on the drive is equal to + /dev/sda1, /dev/sda2, etc, etc. These file can be used for many + things, for example if you want to burn an ISO or IMG file to a + dvd/cd, the dvd will show up as /dev/sdb (or somethimes /dev/sr0 + depending on weather you are using linux, bsd or something + else), and you can use the dd command to copy it to the disk. + You can also use the mount command to access the files on a + drive. You can say mount /dev/sdb1 /mnt (which could be the + first partion of a usb drive currently connected). This will + mean that if you cd into /mnt, you will see the contents of the + usb drive, and you can copy or edit files on /mnt to copy or + edit the content of the drive. +\end{description} + +\subsection{Some examples of things and how to try them} + +\section{The C wizard handbook} +\subsection{Why is C the way it is?} +} +\end{document} |