What is an Operating System?




There are several major core components of an operating system:

The book is divided up into major chapters which deal exhaustively with each topic, but for the purposes of this synopsis, we will deal with these four topics briefly under the headings:

What is an Operating System?

Tanenbaum recommends defining an operating system both:

  1. An “Extended Machine”; and
  2. A Resource Manager.

The best elucidation of what is meant by these two terms comes from the summary of the chapter, because much of these introductory paragraphs are just platitudes that list the components of a computer, and don’t actually explain anything.

From the summary:

With respect to the “resource manager” attribute or aspect of the definition of an operating system, this means that “the operating system needs to efficiently manage the resources of the system”.

Regarding the “extended machine” view, Tenanbaum says: “The job of the operating system is to provide its users with a virtual machine that is more convenient to use than the actual bare metal of the computer”.


A program in execution within an operating system is defined as a “process”.

Tenanbaum: “Associated with each process is its address space, a list of memory locations from some minimum to some maximum, to which the process is permitted to read and write.”

The address space contains the executable program, the program’s data, and its stack. Associated with the program is some set of CPU registers, including the Program Counter (PC), Stack Pointer (SP), &c.


One important key concept of the Process Management aspect of operating systems is “multiprogramming”. This means most often systematically, and at other times and arbitrarily and reactively being able to start, stop, suspend and kill processes at will.

Suspending processes

When a process is suspended, it has to be started in exactly the same state as it was stopped. All the information abut the state of that process has to be recorded down, somewhere, somehow.

In many operating systems, all the information about the process, other than the contents of its own address space, is stored in an operating system table called the process table, which is an array (or linked list) of structures, one for each process currently in existence.

Therefore, a suspended process consists of its address space, usually called its “core image”, alongside its “process table entry”, which contants the state and name of its associated CPU registers.

Process Management System Calls

Commands or requests sent to the operating system to manipulate or read information about processes, whether they be from a process about itself or another one, is called “process management”, and the request or instruction sent to the operating system is called a “system call”.

Even Graphical User Interfaces can manage operating system processes. GUIs are just very simple command interpreters. And of course command line shells or terminal emulators are obviously able to manipulate operating system process management by issuing system calls.

Interprocess Communication


Sometimes we need to convey information to a process that is not waiting or expecting this information. Here’s an example:

A process is sending some information to another host somewhere outside this operating system. The process will request the operating system to notify it of the exhaustion of a time limit by setting a timer, so that the process can determine whether to retransmit the information.

This is called an ‘alarm’ signal. Signals are special instructions that operating systems send to processes to interrupt their normal and autonomous operation. This is a really important and powerful feature of operating systems. Signals are the software equivalent or analogue of hardware interrupts. They are instrumental in being able to negotiate multiprogramming operating system environments.

When a process receives a signal, the process stops what it is doing, saves its state to memory, and goes into a special signal handling procedure. When that is done, the process recovers its old state, and goes back to what it was doing.

There are many and various types of signals, and reasons for using them. (Like the KILL signal that we’re all familiar with >:3). Some interesting signals are ones generated and sent due to traps in the kernel detected by hardware, like illegal instructions or invalid addresses.

I am going to skip over the text on UIDs, GIDs, because everyone I am addressing this to has some conception of UNIX PIDs which is of a good working standard.


Definiton of Some Terms

A “file pathname” is the location of a file relative to the location of the root of the filesystem. File systems in this book are assumed to be hierarchical. So, we all know /home/USERNAME/… etc, or /dev/sda… etc.

Every process has a “working directory”, which is the files in a path relative to this defined path that the process “looks for”.

“File permissions” -> self explanatory, already prior knowledge of most UNIX-like operating system users. Turns out they’re a part of how the operating system manages system calls, processes, and memory!! Who knew??

“Mounting file systems” -> attaching different directory hierarchies, sometimes with different filesystem architectures, to some location relative to the root of some other file system hierarchy.

General Concept of ‘Files’ in Operating Systems

Operating systems provide programmers and users with abstractions of the I/O and disk hardware. That way, the arbitariness and pecularity of the hardware, and the file system digital structure, is hidden, but still able to be manipulated just as effectively, if not more powerfully.

Files, File Systems, and System Calls

System calls, much like in process management, are used to manipulate files and file systems. System calls can:

A lot of these operations are dependent on another concept called ‘opening’ and ‘closing’ files. This is fairly intuitive. A file that hasn’t been opened cannot be written to, and a file that has not been closed has not entered a state in which it is ready to be read by other processes.

It goes without saying that, yep, you guessed it, system calls all deal with manipulating directories of files as well as just the files.

A final note: both processes and file systems are organised hierarchically, but the similarity stops there. Processes usually only exist temporarily, and for matters of minutes, but files have been known to exist in wild for many multiples of years.

Special Files

These objects in file systems are provided in order for devices, and other kinds of complex hardware components to look like files. This special kind of object allows the use of identical system calls for access and manipulation of the hardware.

Special Files in MINIX 3

In MINIX 3, there are two kinds of special files, and as far as I can tell, people who are UNIX-like users will already be very familiar with them:

The first treat randomly accessible hardware components, such as hard disks or CD-ROMs &c, as files in the file system. The great power of this form of hardware control allows one to open up blocks of data on a hardware component through processes without needing to know anything about the structure of the way the files are organised on that opened block at all.

The second is very simple: a character file allows the interfacing of the operating system with hardware that accepts character I/O, say… serial modems or old-timey printers, routers… &c


Pipes allow processes to send information to one-another by tricking them into thinkign that they are writing to files, when in fact the operating system is directly ‘piping’ the information between them.

[A] === [B]

These two processes, for example, think they are writing out their output to regular files in the file system, but actually they are directly transmitting information between themselves. The processes would have to submit a system call in order to discover this greater amount of contextual information about how they are operating.

The Shell

It is, in fact, not a part of the operating system at all. It does, make heavy use of a great number of the features of an operating system.

Comment from Vidak: ((It is also the objectively correct way to interface with a computer system. If you disagree you are wrong and should feel bad. Go home and rethink your life C; ))

System Calls

Tanenbaum: “It is useful to remember: single CPUs can only execute one instruction at a time. If a process is running a user program in user mode and needs a system service, such as reading data from a file, it has to execute a trap or system call instruction to transfer control to the operating system in order to complete this task.

The operating system then inspects the parameters of the calling process in order to work out what it wants.

Speaking broadly and generally, this is the process through which a process makes a system call to the operating system:

MINIX System Calls

As a general rule, MINIX is highly conformant with POSIX standards. However, the mapping of MINIX system calls onto POSIX procedure calls is not always necessarily 1:1. The POSIX standard specifies a number of procedures that a conformant system must supply, but it does not specify/require them to be system calls, library calls, or something else.

In some cases, the POSIX procedures are supported as library routines in MINIX 3. In roders, several required procedures are only minor variations of one-another, and one system call handles all of them.

An Exhaustive List of MINIX 3’s System Calls

There are exactly 53 fundamental system calls in the MINIX operating system. They are divided up into exactly six categories. These are the six categories:

Comment from Vidak:

My decision to include a complete list of all the basic system calls in MINIX that are written in Chapter One is because having some familiarity with any or all of them is very enlightening and informative. Each of these system calls is conformant with some critical requirement in the POSIX standard. This means just reading the names and descriptions of each of the system calls helps you learn the POSIX standard.

As far as I am aware, a complete operating system, even if just a toy one, can be assembled by implementing all of these system calls. This is a very exciting fact to know. It means you can have a powerful working knowledge of operating systems just by knowing this small number of system calls (what is elsewhere called an API, in proprietary operating system documentation).

To better aid memorisation or quick mental recall, I have added mneumonics next to each brief system call definition. The structure of the sections that follow is the following:



(Number of system calls: 10)

0: pid = fork()

(“FORK”) Create a child process identical to the parent.

This is the only way to make a new process in MINIX 3. It creates an exact duplicate of the original process, including all the file descriptors, registers – /everything/. After the fork, the original process (parent) and the copy (child) go their separate ways. All the variables have identical values at the time of the fork, but since the parent’s data are copied to create the child, subsequent changes in one of them do nto affect the other – however, at this stage, the program text of both the parent and child are identical.

The fork call returns a value which is zero in the child and equal to the child’s PID in the parent. Using the returned PID, the two processes can see which one is the parent, and which one is the child.

1: pid = waitpid(pid, &statloc, opts)

(“WAITPID”) Wait for a child process to terminate.

In most cases, the child will need to execute different code than the parent. This is where this system call is very important. This system call is an instruction that the parent process sends to the operating system. Its purpose is to request the operating system to suspend the parent in order to wait for the completion of the execution of the child process.

The address pointed to by the second parameter, statloc, will be set to the child’s exit status. The third parameter offers various other options, and makes a little interesting extra reading.

2: s = wait(&status)

(“WAIT”) Old version of waitpid(…)

3: s = execve(name, argv, envp)

(“EXECVE”) Replace the entire core image of a process.

To affect some interesting consequence in the world, the child process is somehow going to have to change its program code. This system call achieves that purpose. It replaces the entire core image of the child process. The new core image that the child process receives is the first parameter of the system call, ‘name’.

This is only the basic gist of how this flow works, however, because, in MINIX it is actually a little more complex. The actual system call that is performed to replace a process’s core image is called ‘exec’, and the system call written here is actually one of the subsequent library calls that exec makes.

The extra complexity on this point is actually very good – it means you have many different parameters and options to consider when forking.

This is a non-exhaustive list of the different library routines which assist the primitive exec system call:

These different libraries are provided to allow the parameters of exec to be omitted or specified in various ways. Generally, throughout the MINX 3 book, these are all grouped together under the abstract concept of exec.

In the msot general case of forking, execve has three parameters:


Consider a practical example. How does the operating system execute the following shell command?

cp file1 file2

The program code of cp is, lamentably, written in C, but its main procedure looks like this:

main(argc, argv, envp)

So, when loading the cp program data into a child process’s core image, the operating system will fill in all these variables for the child process, so that it can execute the code, complete its program, and then exit. All of this was triggered by the parent process’s request to the operating system for a forking of its process, according to the parameters it passed along with the FORK system call.

What happens after the child process has completed the execution of its program?

4: exit(status)

(“EXIT”) Terminate the execution of a process and return an exit status number.

The child process will pass an EXIT system call to the operating system. This system call will pass a status value with its call, which, in MINIX (and, I think this is POSIX conformant, so therefore true for all UNIX operating systems), is number from 0 to 255 (i.e., $0000 to $FF - I refuse to use the notation ‘0xFF’).

This exit status is actually passed to the parent process. It is passed into the &statloc argument of its completed WAITPID system call (see above).

If I am correct, &statloc is a 16-bit argument. The lower 8-bits of the argument contains the termination status, with 0 being normal termination and the other values being various error conditions. The higher 8-bits contains the child’s exit status.

Say, if a child process exits its execution with ‘4’ as the parameter to exit, the parent will be awakened by the operating system with n set to the child’s PID, and &statloc set to the constant $0400.

5: size = brk(addr)

(“SIZE”/“BRK”) Set the size of the process’s data segment (i.e., where the process’s variables are stored in RAM).

Now we can move onto a different topic: the anatomy of processes in MINIX and POSIX systems. In MINIX, processes have their memory divided up into three segments:

Speaking generally, this illustration represent the memory map of a typical process:

STACK (grows downwards)
DATA (grows upwards)

As the process moves through execution, the data segment grows ever upwards into memory, and the stack grows ever downwards. Between them is unused memory/address space.

The stack grows into the gap automatically, as needed. The expansion of the data segment, however, is done explicitly through the use a system called BRK (the title of this heading) which specifies the new address where the data segment is to end.

This address may be more or less than the current value, however it is forbidden to have the data segment become more than the stack.

BRK, and SBRK (the semi-automatic library for the BRK system call), are however, NOT POSIX.

Usually, programmers, at time of this writing of the MINIX book (2008) are encouraged to use malloc. Hilariously, malloc is not specified in POSIX, because it was thought not necessary.

6: pid = getpid()

(“GETPID”) Return caller’s PID.

7: pid = getpgrp()

(“GETPGRP”) Return ID of caller’s process group.

8: pid = setsid()

(“SETSID”) Create new ‘session’ and return its process group ID.

9: l = ptrace(req, pid, addr, data)

(“PTRACE”) Used for debugging. There is a shell command called ptrace. Running it yourself is completely safe, and it’s a great way to learn about how all this stuff works! :-)


(Number of system calls: 8)

When a signal is sent to a process that has not announced its willingness to accept that signal, the process is unilaterally ‘killed’. To avoid this fate, a process can use the SIGACTION system call:

0: s = sigaction(sig, &act, &oldact)

(“SIGACTION”) Define the action that is to be taken on a transmitted signal.

With this system call, a process can announce that it is prepared to accept some signal type, and to provide the address of the signal handling procedure and a place to store the address of the current signal.

After a SIGACTION call, if a signal of the relevant type is generated, the state of the process is pushed onto its own stack, and then the signal handler is called. It may run for as long as it wants to, and perform any system calls it wants.

When the signal handling is done, it calls SIGRETURN (immediately below) to continue where it left off.

1: s = sigreturn(&context)

(“SIGRETURN”) Return from a signal.

2: s = sigprocmask(how, &set, &old)

(“SIGPROCMASK”) Examine or change the signal mask.

Signals can be blocked, by processes, but not indefinitely. ‘Blocked’ signals are held pending until unblocked. This system call, SIGPROCMASK, examines or manipulates a bitmap of ‘a’ or ‘the’ set of current blocked system signals.

3: s = sigpending(set)

(“SIGPENDING”) Find out the set of signals that are currently blocked.

A process can use this system call to inquire and learn about the pending set of signals that are currently undeliverable due to being blocked. This information is returned as a bitmap.

4: s = sigsuspend(sigmask)

(“SIGSUSPEND”) Replace the signal mask and suspend the selected process.

Using this system call, a process may suspend itself by setting a bitmap of suspended system processes.

5: s = kill(pid, sig)

(“KILL”) Send a signal to a process.

A signal can prevent subsequent types of signals from the operating system with SIG_IGN, and can return to default action responses to signals with SIG_DFL.

Vidak: ((I assume both of these variables manipulate the system bitmap of signals.))

However, the default action is either to KILL a process or ignore the signal, depending on which signal it is, in the current context.

Critically, there are virtually no processes can can ignore or capture and ‘block’ the signal KILL when it is exercised unilaterally by the operating system.

6: residual = alarm(seconds)

(“RESIDUAL”) Set the alarm clock.

7: s = pause()

(“PAUSE”) Suspend the caller until the next signal.


(Number of system calls: 15)

0: fd = creat(name, mode)

(“CREAT”) Obsolete way to create a new file.

1: fd = mknod(name, mode, addr)

(“MKNOD”) Create regular, special, or directory i-node.

2: fd = open(file, how, …)

(“OPEN”) Open file for reading, writing, or both.

3: s = close(fd)

(“CLOSE”) Close an open file.

4: n = read(fd, buffer, nbytes)

(“READ”) Read data from file into buffer.

5: n = write(fd, buffer, nbytes)

(“WRITE”) Write data from buffer into file.

6: pos = lseek(fd, offset, whence)

(“LSEEK”) Move the file pointer information. (So, we’re now pointing somewhere else).

7: s = stat(name, &buff)

(“STAT”) Get the file’s status.

8: s = fstat(fd, &buf)

(“FSTAT”) Get file’s status information.

9: fd = dup(fd)

(“DUP”) Allocate new file descriptor for open file.

10: s = pipe(&fd[0])

(“PIPE”) Create pipe.

11: s = ioctl(fd, request, argp)

(“IOCTL”) Perform special operations on a file.

12: s = access(name, amode)

(“ACCESS”) Check a file’s accessibility.

13: s = rename(old, new)

(“RENAME”) Give a file a new name.

14: s = fcntl(fd, cmd, …)

(“FCNTL”) Lock a file, and other sorts of operations.


(Number of system calls: 9)

These system calls your average UNIX user would use perhaps 200 times a week, so I will not spend time explaining them, as they are really a part of basic UNIX literacy.

0: s = mkdir(name, mode)

1: s = rmdir(name)

2: s = link(name, name2)

3: s = unlink(name)

4: s = mount(special, name, flag)

5: s = umount(special)

6: s = sync()

7: s = chdir(dirname)

8: s = chroot(dirname)


(Number of system calls: 7)

Again, all of this is a part of basic UNIX literacy. It is quite happy that basic UNIX (shell) literacy is virtually immediate instruction and familiarity with directly controlling the system kernel. There are no intermediate metaphors, like the ‘app store’ or ‘siri’/‘cortana’.

0: s = chmod

1: uid = getuid()

2: gid = setgid()

3: s = setuid(uid)

4: s = setgid(gid)

5: s = chown(name, owner, group)

6: oldmask = umask(complmode)


(Number of system calls: 4)

0: seconds = time(&seconds)

(Get elapsed time since 1970-01-01).

1: s = stime(tp)

(Set elapsed time since 1970-01-01).

2: s = utime(file, timep)

(Set file’s ‘last access’ time).

3: s = times(buffer)

(Get the user and system times used so far).

Conclusion: Monolithic Operating Systems

Tanenbaum argues that monolithic operating systems effectively have no structure.

These operating systems are an enormous bank which is a collection of all the procedures that form the entire operating system.

Any procedure may call any other. The thing that does organise and structure the monolithic operating system is that each procedure has well defined interfaces, in terms of their paramters and expected results.

“To construct the actual object project of the monolithic operating system when this approach is used, one:

(1) First compiles all the individual procedures (or files containing the procedures) (2) Then binds them together into a single object file using the "system linker”

Monolithic kernels have essentially no “information binding”. Every procedure is visible to every other procedure. This is counterposed to a structure where an operating system is organised into modules or packages. This alternative structure has designated entry and exit points for each module, and will otherwise conceal the rest of the information about its operation.

Monolithic systems can have a little structure: the “services” or, more POSIX-like, “system calls” are all provided and implemented by structuring them in such a way that their paramters are placed in well-known locations like, (a) certain registers; (b) the stack.

In monolithic systems, the supervisor call switches the computer from user to kernel mode, and transfers control the user to the operating system. Most CPUs are structured in this way: (userland/superuser space).

An Example System Call

Let us take this example system call, to illustrate this switching between usermode and superuser mode:

count = read(fd, buffer, nbytes);

  1. Push parameters nbytes, buffer, and fd onto the stack, in that order – the reverse order. This is done by convention in C and C++ because it is convenient and desirable to make the first parameter to the format string PRINTF appear at the top of the stack. ((Don’t worry, I don’t understand either.))

    The parameters fd and nbytes are called by value, but buffer is caleld by reference – which means the address of the buffer is passed, and not the actual value of the contents of the buffer.

  2. We then call the library procedures which prepare the data for calling the operating system. This instruction is the normal procedure call instruction that we use to call all procedures in MINIX. The library is possibly written in ASM. The library puts the system call numebr in a place where the operating system expects it (for example, in some CPU register designated by convention).

  3. The TRAP instruction is then executed, and we switch from user mode to superuser mode. Execution is usually started a fixed address.

  4. The kernel takes the system call binary or hex value, and examines it, and then dispatches it to the system call handler. Usually this is done with a system of pointers to tables that would call the handlers of this particular system call number input.

  5. The system call handler is selected by the pointer tables, and runs.

  6. The system call handler finishes its work, and CPU control is now handed back to userspace library procedure that called this whole process. Execution is resumed at the exact instruction after the TRAP command.

  7. Return is called, and we’re now back in the userland program that called the system call.

  8. The stack now needs to be cleaned up. The quick and dirty way to do this is usually to point the Stack Pointer where it will overwrite the old kernel system call and system call argument values.

Tanenbaum’s Suggested Monolithic Operating System Structure

  1. A main userland program which invoked or executes the system call procedure.
  2. A set of service procedures which carry out the system calls.
  3. A set of utility procedures which help the system call procedures.

For each system call you’d ideally have one service procedure. The utility procedures then do things that many service procedures require, like getting the right data from userland.




1 174
16-bit 1
1984 1
2009 1
3d-graphics 1
400 1
60s 1
6502 1
8-bit 1
8-bit-cpu 1
abc-news 1
absolutism 1
abstraction 1
activism 2
addicted 1
advice 1
age 1
agency 1
aging 1
alasdair-macintyre 1
alienation 1
almost-over 1
alt-right 1
america 1
anaesthetists 1
anarchism 12
anarchist 2
ancap 1
ancom 1
anime 3
anthropic 1
anthropic-principle 1
anthropy 1
anti-authoritarian 2
anti-essentialism 1
antichamber 1
antics 2
aoc 1
aphorisms 1
aretha 1
aristotle 1
atari 1
atari-2600 2
atheism 1
attitude-changes 1
auspol 1
australia 4
australian-government 1
australian-politics 1
awesome 2
bad 2
banjo-kazooie 1
base 1
basics 1
battlefield-bad-company 1
be-my-friend 1
beach 1
beatles 1
being-myself 1
bernie-sanders 1
binary-logic 1
biography 1
bioshock 1
bivariate-statistics 1
black-bloc 1
blah 1
blair-leeching-off-everyone 1
blender 1
blog 1
blow 1
bones 1
book 1
bookchin 1
books 1
bootloader 1
brain-surgery 1
breakfast 2
bullying 2
bureaucracy 1
burnt 1
burqa 1
bus 1
but-i-have-no-food 1
c 3
c-programming 1
calculus 1
calm 1
camera 1
can039t-sleep 3
canteen 1
capitalism 11
cat 3
cat-hair 1
cats 1
centrelink 1
chances 1
charter 1
chat-bot 1
chatswood 1
checkout-chick 1
checkouts 1
chemistry 1
chemistry-essay 1
chicken 1
children 1
chocolate 1
choice 1
chomksy 1
civilisation 1
code-46 1
coding 9
coffee 1
coffee-review 1
cold-war 1
coles 1
collapse 1
colonialism 1
commodities 1
common-lisp 1
communism 2
communist 3
communist-economics 1
community 1
compilers 1
computer-architecture 1
computer-science 2
computers 2
concepts 1
consumerism 1
content 1
continent 1
convention 1
cool 2
core-basics 1
cough-mixture 1
council 1
covid 1
covid-19 4
cpu 2
creative 1
crisis 1
criticism 1
cryptic-shit 1
culture 1
cyber-bullying 1
cybernetics 1
cybersecurity 1
cyberspace 3
darkness 1
debian 1
debut 1
democracy 1
depression 2
desert 1
dialectics 3
direct-action 1
disagreement 1
disco 1
discrete-transistors 1
disney 1
doctor 1
doctors 1
dog 1
domain-name 1
donald-trump 1
dont-hate-me 1
down-to-clown 1
dr 1
draft 1
drawing 1
dreamcast 3
drm 1
druggies 1
druggies-on-the-bus 2
drugs 2
dtl-logic 1
dusty 1
easy-way-through-life 1
eat 1
economics 5
economy 1
edward-said 1
egotism 1
election 1
elections 1
electronics 2
emacs 3
emotivism 1
engels 1
engineering 1
english 1
entertainment 1
epic-poems 1
epicurics 1
epicurus 1
espresso 1
essay 2
essential-left 1
essentialism 1
ethical-substance 1
ethics 3
evangelion 1
everything 3
everythings-sorted-up-here 1
evo-psych 1
evolution 1
evolutionary-psychology 1
ex-lj 13
exchange 2
expensive 1
facebook 1
failure 1
fair-work-commission 1
fairy-tales 1
false-rumours 1
famiily 1
families 1
family 1
fascism 1
fast-food 1
fat 1
fat-people 1
fear 1
fediverse 5
feuerbach 3
fiction 1
film 1
final-fantasy 1
fiscal 1
floreat 1
floreat-library 1
folk-tales-from-malasrion 2
food 2
food-review 1
form 1
forth 1
foss 16
free-markets 1
free-software 1
freedom 2
friends 1
fun 1
fun-in-the-floreat-library 1
functional-programming 1
fundamentals 1
funk 1
funny 10
funny-shit 1
future 1
game 1
game-dev 1
game-gear 1
gamecube 1
gamergate 1
games 2
gaming 5
gatherings 1
gemini 1
gemini-protocol 1
genesis 1
give-back 1
give-me-money 1
give-you-up 1
global 1
global-warming 1
globe 1
gnu 3
god 2
gold-standard 1
google 1
graph-paper 1
graphic-novel 1
graveyard 1
greens 1
gremanese 1
gremano 5
growth 1
guix 6
guns 1
habits 1
hackaday 1
hacker 1
hacking 6
hacks 1
hail 1
hair 1
hale 1
hale-school 1
half-life-2 1
half-light 1
happiness 3
happy 1
hardware 7
hasbro 1
hatespeech 1
hegel 8
hermeneutics 2
highly-online 1
hijab 1
hilarious 1
hip 1
hippie-shit 1
history-of-computers 1
hobbies 1
hobby 1
holidays 1
hollywood 1
homebrew-computer 1
homebrew-cpu 1
hope 1
hospital 3
hotel 1
houses 1
housing 1
how-to 1
html 1
http 1
hugs 1
humour 6
hurt 2
i-am 1
i-am-going-to-waste-money-on-making-a-website 1
i-am-hilarious 1
i-am-making-a-website 1
i-dont-have-any-phone-credit-left 1
i-have-no-food 1
i-herd-u-liek-ruddkips 1
i-want-food 1
i-want-to-work-in-a-bookstore 1
i-wrote-this-a-while-ago 1
ideology 1
immaturity 1
imperialism 1
indie-project 2
indigenous-cultures 1
individualism 1
industries 1
inflation 1
insanity 1
institutions 1
insurrection 1
interdimensional 1
interesting 1
interpersonal-politics 2
interpretation 1
interpreters 1
intrinsic 1
investment 2
invisible-hand 1
iron 1
iron-man 1
irony 1
islam 1
islamophobia 1
it-can-be-your-website-too 1
japan 2
japanese 2
javascript 1
jeff 1
jen 1
jewish-question 1
jjjm 1
joe 1
john-lennon 1
joke 4
jokes 1
jon-blow 1
journalism 2
jousen 1
jumper 1
just-give-up-now 1
justice 2
juvenile 1
k-and-r-ansi-c 1
kafkaesque 1
kantianism 1
kernels 1
kindle 1
king 1
kropotkin 1
labour-theory-of-value 1
lack-of-family 1
lack-of-sleep 1
landships 1
language 1
large-men 1
leadership 1
leavers 1
leederville 1
leftist 1
legacy-cpu 1
lenin 1
leonid-brezhnev 1
lesser-jou 1
letters 1
letters-that-tell-stories 1
liberalism 1
life 2
lifestyle 1
lights-on-the-horizon 1
lineageos 1
linux 11
lisp 12
lisp-machine 1
literature 2
lobotomy-software 2
lockdown 2
logic 1
looking-to-the-past 1
loser 1
loss 1
love 3
lucretius 1
machiavelli 1
macintyre 1
macros 1
mainstream-media 1
malasrion 21
man 1
markdown 1
markets 2
marx 7
marxism 2
marxist-economics 1
master-system 1
mastodon 1
materialism 1
mathematics 2
maths 1
matt 1
may-day 1
maybe-in-elizabeths 1
mcdonalds 1
mcdowell 1
meaning 2
mechanical-marxism 1
media 2
mega-drive 10
memes 1
memories 1
mental-health 2
metaphysics 1
methodone 1
micro-kernels 2
middle-east 1
midnight 1
mightve-spelt-methodone-wrong 1
minecraft 1
minix 2
mmorpg 1
money 2
monies 1
moon 1
morality 3
morals 1
more-stories-from-malasrion 1
movements 1
movie 4
movies 2
mozilla 1
mr-shaw 1
mua-union 1
music 1
musings 1
muslim 1
mutually-assured-destruction 1
n64 11
natural 1
naturalism 1
navel-gazing 1
nazi 1
nazism 1
need-help 1
nela 1
neo-nazi 1
neolithic 1
nes 7
netflix 1
never-gonna-give-you-up 1
newgrounds-flash 1
night 3
nightfill 1
nihilism 1
nintend-wii 1
nintendo 5
no-left-turn 1
nonsense 1
normativity 1
norms 1
north-perth 1
not-boring 1
nuclear-explosions 1
nuclear-family 1
obesity 1
objectivism 1
observations 1
off-the-battlefield 1
office 1
olol 1
on-the-battlefield 1
online 1
online-communities 1
online-culture 2
online-history 1
open-hardware 3
open-heart-surgery 1
operating-system 2
operating-systems 2
oppression 1
organicisim 1
organising 1
orientalism 1
outline-for-a-bit 1
outrageous 1
over-production 1
pacifism 1
packages 1
pancakes 1
panic 1
pants 1
parliament 3
pc 6
pc-engine 1
pc-fx 1
peace 1
peaceful 1
pen 1
people 1
people-always-text-me-when-i-have-no-phone-credit 1
personal 2
perth 4
pet 1
peter-goldsworthy 1
pets 1
phd 1
phenomenology 1
philosophy 14
philosophy-of-computers 1
philosphy-of-computers 1
phone-unlock 1
pictures 1
pills 1
pine64 1
pinephone 3
plains 1
plan-for-game 1
plan9 1
platform 1
playstation 1
please-dont-sue-me 1
pleroma 1
poems 1
poetry 3
pokemon 1
pokemon-clone 1
polemic 1
political 1
political-economy 1
politics 16
polygonz 1
pop 1
postgrad 1
poverty 1
precarious-work 1
price 1
price-and-profit 1
prisoners-dilemma 1
privacy 1
private-property 1
prizes 1
probably-my-shittest-poem-yet 1
productivity 1
productivity-commission 1
profit 2
programming 2
project 3
projects 3
proposal 1
proprietary 3
proudhon 2
ps1 2
ps2 2
ps3 2
public 1
public-office 1
public-transport 1
punching-richard-spencer 1
qantas 1
quake 1
quake-gl 1
quietism 1
racism 2
radical 1
radio 1
radiohead 1
rain 1
ranciere 1
random-updates 19
rare 2
raspberry-pi 1
rational 1
rational-functions 1
rationalism 1
realistic-space 1
reason 1
red-and-sven 2
relativism 1
representation 1
representative 1
resolutions 1
retail 1
retro 1
retro-games 1
review 2
revolution 1
rich-kids 1
riches 1
rick 1
rick-astley 1
right 1
rightist 1
road-to-no-where 1
robot 1
robot-cat-freaked-me-out 1
robots-are-real 1
robotssss 1
roguelike 1
romantic 1
romanticism 1
rottnest 1
rudd 1
ruddkips 1
russia 1
sackings 1
salary 1
sales 1
saturday 2
saturn 1
scam 1
school 3
science 1
scomo 1
screaming-children 1
screens 1
searching 1
secure-work 1
sega 3
sexism 1
shares 1
shelf-stacking 1
shitty 1
shitty-long-poem 1
shrendig-and-foolio 1
sieg-heil 1
singapore 2
single-board-computers 1
sitcom 1
sittlichkeit 1
sleep 2
sly-and-the-family 1
smolnet 1
snes 7
socdem 1
social 2
social-democracy 1
social-media 4
social-problems 1
socialism 3
socialist 1
socialist-strategy 1
socially-necessary-labour-time 1
society 2
software-hacking 1
someone-give-me-food 1
sony 2
soul 1
soviet 2
soviets 1
spin 1
spinoza 1
spoof 1
square-enix 1
ssh 1
star-fox 1
star-fox-adventures 1
star-trek 1
states 2
stimulus-package 1
stoicism 1
stone 1
strategy 1
study 1
stuff 4
stupidity 1
sun 1
sunday 1
sunset 1
sunshine-on-the-spotless-mind 1
superstructure 1
supremacy 1
surgery 1
surplus-value 2
surveillance 1
tactics 1
tamagotchi 3
tanjoubi-omedetou 1
tankie 1
tech 9
technology 2
teeth 1
teleology 1
television 1
ternary-logic 1
terrorism 1
text 1
text-files 1
the-almighty-flarg 1
the-king 1
the-notion 1
theft 3
theory 5
therapeutic 1
therapy 1
thesis 1
things 1
things-that-are-burnt 1
things-that-do-stuff 1
tic-80 1
ticket 1
tickets 1
time-travel 1
toys 1
trackers 1
traditional-saying 1
transactions 1
transistors 1
trendy 1
trotskyism 1
trust 1
ttl-logic 1
tuck-shop 1
turbografx 1
two-suns 1
ubuntu 1
ugliness 1
ugly-people 3
uncategorised 59
uncategorized 4
uni 2
union-unrest 1
unionism 1
unions 2
united-states 1
university 1
unix 4
unix-philosophy 1
ussr 1
value 1
values 1
victim 1
video-game 1
video-games 3
violence 2
virtual-pet 1
virtue-ethics 1
w-b-yeats 1
wages 1
wai-con 1
wai.con 1
walking 1
war 2
wealth 1
web 3
website 1
weird 2
weirdness 2
wembley-downs 1
western-australia 1
western-civilisation 1
where-is-the-battlefield 1
whine 1
whiney-people 1
white 1
white-supremacist 1
wifi 2
wii-u 2
william-butler-yeats 1
windows 1
woo 1
woolworths 1
work 6
work-is-shit 1
worker-unrest 1
workers 3
workers-liberty 1
workers-rights 1
world-wide-web 1
xbox 1
xiaomi 1
yeats 1
young-people 2
young-worker 1
youth 1
zelda 1
zizek 1