In general *bootstrapping* (from the idiom "pull yourself up by your bootstraps"), sometimes shortened to just *booting*, refers to a clever process of automatically self-establishing a relatively complex system starting from something very small, without much external help. Nature itself provides a beautiful example: a large plant capable of complex behavior (such as reproduction) initially grows ("bootstraps") from just a very tiny seed. As another example imagine something like a "civilization bootstrapping kit" that contains only a few primitive tools along with instructions on how to use those tools to mine ore, turn it into metal out of which one makes more tools which will be used to obtain more material and so on up until having basically all modern technology and factories set up in relatively short time ([civboot](civboot.md) is a [project](project.md) like this). The term *bootstrapping* is however especially relevant in relation to [computer](computer.md) technology -- here it possesses two main meanings:
- The process by which a computer starts and sets up the [operating system](operating_system.md) after power on, which often involves several stages of loading various modules, running several bootloaders etc. This is traditionally called **booting** (*rebooting* means restarting the computer).
- Utilizing the principle of bootstrapping for making greatly independent [software](software.md), i.e. software that doesn't [depend](depend.md) on other software as it can set itself up. This is usually what **bootstrapping** (the longer term) means. This is also greatly related to **[self hosting](self_hosting.md)**, another principle whose idea is to "implement technology using itself".
Bootstrapping -- as the general concept of letting a big thing grow out of a small seed -- may aid us in building extremely [free](free_software.md) (as in freedom), [portable](portability.md), self-contained (and yes, for those who care also more [secure](security.md)) technology by reducing all its [dependencies](dependency.md) to a bare minimum. If we are building a big computing environment (such as an operating system), we should make sure that all the big things it contains are made only with the smaller things that are further on built using yet smaller things and so on until some very tiny piece of code, i.e. we shall make sure there is always a way to set this whole system from the ground up, from a very small amount of initial code/tools. Being able to do this means our system is *bootstrappable* and it will allow us for example to set our whole system up on a completely new computing platform (e.g. a new CPU architecture) as long as we can set up that tiny initial prerequisite code. This furthermore removes the danger of dependencies that might kill our system and also allows security freaks to inspect the whole process of the system set up so that they can trust it (because even free software that sometime in the past touched a proprietary compiler can't generally be trusted -- see [trusting trust](trusting_trust.md)). I.e. bootstrapping means creating a very small amount of code that will self-establish our whole computing environment by first compiling small compilers that will then compile more complex compilers which will compile all the tools and programs etc. This topic is discussed for example in designing [programming language](programming_language.md) [compilers](compiler.md) and [operating systems](os.md). For examples of bootstrapping see e.g. [DuskOS](duskos.md) ([collapse](collapse.md)-ready operating system that bootstraps itself from a tiny amount of code), [GNU](gnu.md) [mes](mes.md) (bootstrapping system of the GNU operating system) or [comun](comun.md) (LRS programming language, now self hosted and bootstrappable e.g. from a few hundred lines of [C](c.md)).
**Why concern ourselves with bootstrapping when we already have our systems set up?** Besides the obvious elegance of this whole approach there are many other practical reasons -- as mentioned, some are concerned about "security", some want portability, control and independence -- one of other notable justifications is that we may lose our current technology due to societal [collapse](collapse.md), which is not improbable as it keeps happening throughout history over and over, so many people fear (rightfully so) that if by whatever disaster we lose our current computers, Internet etc., we will also lose with it all modern art, data, software we so painfully developed, digitized books and so on; not talking about the horrors that will follow if we're unable to quickly reestablish our computer networks we are so dependent on. Setting up what we currently have completely from scratch would be extremely difficult, a task for centuries -- just take a while to consider all the activity and knowledge that's required around the globe to create a single computer with all its billions of lines of code worth of software that makes it work. Knowledge of old technology gets lost -- to make modern computers we first needed older, primitive computers, but now that we only have modern computers no one remembers anymore how to make the older computers -- modern computers are sustaining themselves but once they're gone, we won't know how to make them again, i.e. if we lose computers, we will also lose tools for making computers. This applies on many levels (hardware, operating systems, programming languages and so on).
Bootstrapping has to start with some initial prerequisite machine dependent binary code that kickstarts the self-establishing process, i.e. it's not possible to get rid of absolutely ALL binary code and have a pure bootstrappable code that would run on every computer -- that would require making a program that can native run on any computer, which can't be done -- but it is possible to get it to absolute minimum -- let's say a few dozen bytes of machine code that can even be hand-made on paper and can be easily inspected for "safety". This initial binary code is called *bootstrapping binary seed*. This code can be as simple as a mere translator of some extremely simple bytecode (that may consist only of handful of instructions) to the platform's assembly language. There even exists the extreme case of a single instruction computer, but in practice it's not necessary to go as far. The initial binary seed may then typically be used to translate a precompiled bytecode of our system's compiler to native runnable code and voila, we can now happily start compiling whatever we want.
[Forth](forth.md) is a language that has traditionally been used for making bootstrapping environments -- its paradigm and philosophy is ideal for bootstrapping as it's based on the concept of building a computing environment practically from nothing just by defining new and new words using previously defined simpler words, fitting the definition of bootstrapping perfectly. [Dusk OS](duskos.md) is a project demonstrating this. Similarly simple language such as [Lisp](lisp.md) and [comun](comun.md) can work too (GNU Mes uses a combination of [Scheme](scheme.md) and C).
1.**Make a [simple](kiss.md) [programming language](programming_language.md) L**. You can choose e.g. the mentioned [Forth](forth.md) but you can even make your own, just remember to keep it extremely simple -- simplicity of the base language is the key feature here. If you also need a more complex language, write it in L. The language L will serve as tool for writing software for your platform, i.e. it will provide some comfort in programming (so that you don't have to write in assembly) but mainly it will be an **[abstraction](abstraction.md) layer** for the programs, it will allow them to run on any hardware/platform. The language therefore has to be **[portable](portability.md)**; it should probably abstracts things like [endianness](byte_sex.md), native integer size, control structures etc., so as to work nicely on all [CPUs](cpu.md), but it also mustn't have too much abstraction (such as [OOP](oop.md)) otherwise it will quickly get complicated. The language can compile e.g. to some kind of very simple [bytecode](bytecode.md) that will be easy to translate to any [assembly](assembly.md). Make the bytecode very simple (and document it well) as its complexity will later on determine the complexity of the bootstrap binary seed. At first you'll have to temporarily implement L in some already existing language, e.g. [C](c.md). NOTE: in theory you could just make bytecode, without making L, and just write your software in that bytecode, but the bytecode has to focus on being simple to translate, i.e. it will probably have few opcodes for example, which will be in conflict with making it at least somewhat comfortable to program on your platform. However one can try to make some compromise and it will save the complexity of translating language to bytecode, so it can be considered ([uxn](uxn.md) seems to be doing this).
2.**Write L in itself, i.e. [self host](self_hosting.md) it**. This means you'll use L to write a [compiler](compiler.md) of L that outputs L's bytecode. Once you do this, you have a completely independent language and can start using it instead of the original compiler of L written in another language. Now compile L with itself -- you'll get the bytecode of L compiler. At this point you can bootstrap L on any platform as long as you can execute the L bytecode on it -- this is why it was crucial to make L and its bytecode very simple. In theory it's enough to just interpret the bytecode but it's better to translate it to the platform's native machine code so that you get maximum efficiency (the nature of bytecode should make it so that it isn't really more diffiult to translate it than to interpret it). If for example you want to bootstrap on an [x86](x86.md) CPU, you'll have to write a program (L compiler [backend](backend.md)) that translates the bytecode to x86 assembly; if we suppose that at the time of bootstrapping you will only have this x86 computer, you will have to write the translator in x86 assembly manually. If your bytecode really is simple and well made, it shouldn't be hard though (you will mostly be replacing your bytecode opcodes with given platform's machine code opcodes). Once you have the x86 backend, you can completely bootstrap L's compiler on any x86 computer.
3.**Further help make L bootstrapable**. This means making it even easier to execute the L bytecode on any given platform -- you may for example write backends (the bytecode translators) for common platforms like x86, ARM, RISC-V, C, Lisp and so on. You can also provide tests that will help check newly written backends for correctness. At this point you have L bootstrappable without any [work](work.md) on the platforms for which you provide backends and on others it will just take a tiny bit of work to write its own translator.
4.**Write everything else in L**. This means writing the platform itself and software such as various tools and libraries. You can potentially even use L to write a higher level language (e.g. C) for yet more comfort in programming. Since everything here is written in L and L can be bootstrapped, everything here can be bootstrapped as well.
However, a possibly even better way may be the [Forth](forth.md)-style **incremental programming** way, which works like this (see also [Macrofucker](macrofucker.md) and [portability](portability.md) for explanation of some of the concepts):
1.**Start with a trivially simple language.** It must be one that's easy to implement from scratch on any computer without any extra tools -- something maybe just a little bit more sophisticated than [Brainfuck](brainfuck.md). This language may even be a machine specific [assembly](assembly.md), let's say [x86](x86.md), that's using just a small subset of the simplest instructions, as long as it's easy to replace these instructions with other instructions on another hardware architecture. There should basically only be as many commands to ensure [Turing Completeness](turing_complete.md) and good performance (i.e. while an increment instruction may be enough for Turing completeness, we should probably also include instruction performing general addition, because adding two numbers in a loop using just the increment instruction would be painfully slow). The goal here is of course to build the foundations for the rest of our platform -- one that's simple enough to be easily replaced.
2.**Build a more complex language on top of it.** I.e. now use this simple language ALONE to build a more complex, practically usable language. Again, take inspiration in Forth -- you may for example introduce something like procedures, [macros](macro.md) or words to your simple language, which will allow you to keep adding new useful things such as arrays or more complex control structures. To add the system of macros for example just write a [preprocessor](preprocessor.md) in the base language that will take the new, macro-enabled language source code and convert it to the plain base language; with macros on your disposal now you can start expanding the language more and more just by writing new macros. I.e. expanding the base language should be done in small steps, incrementally -- that is don't build C out of Brainfuck right away; instead first build just a tiny bit more complex language on top of the initial language, then a bit more complex one on top of that etc. -- in Forth this happens by defining new words and expanding the language's dictionary.
3.**Now build everything else with the complex language.** This is already straightforward (though time consuming). First you may even build more language extensions and development tools like a debugger of [text editor](text_editor.md) for example. The beauty of this approach is really that to allow yourself to program on the system you are building the system itself on-the-go, i.e. you are creating a development environment and also a user environment for yourself, AND everything you make is bootstrappable from the original simple language. This is a very elegant, natural way -- you are setting up a complex system, building a road which is subsequently easy to walk again from the start, i.e. bootstrap. This is probably how it should ideally be done.
Booting as in "staring computer up" is also a kind of setting up a system from the ground up -- we take it for granted but remember it takes some [work](work.md) to get a computer from being powered off and having all RAM empty to having an operating system loaded, hardware checked and initialized, devices mounted etc.
Starting up a **simple computer** -- such as some [MCU](mcu.md)-based [embedded](embedded.md) [open console](open_console.md) that runs [bare metal](bare_metal.md) programs -- isn'tas complicated as booting up a mainstream [PC](pc.md) with an [operating system](operating_system.md).
First let's take a look at the simple computer. It may work e.g. like this: upon start the [CPU](cpu.md) initializes its registers and simply starts executing instructions from some given memory address, let's suppose 0 (you will find this in your CPU's data sheet). Here the memory is often e.g. [flash](flash.md) [ROM](rom.md) to which we can externally upload a program from another computer before we turn the CPU on -- in game consoles this can often be done through [USB](usb.md). So we basically upload the program (e.g. a game) we want to run, turn the console on and it starts running it. However further steps are often added, for example there may really be some small, permanently flashed initial boot program at the initial execution address that will handle some things like initializing hardware (screen, speaker, ...), setting up [interrupts](interrupt.md) and so on (which otherwise would have to always be done by the main program itself) and it can also offer some functionality, for example a simple menu through which the user can select to actually load a program from SD card to flash memory (thanks to which we won't need external computer to reload programs). In this case we won't be uploading our main program to the initial execution address but rather somewhere else -- the initial bootloader will jump to this address once it's done its work.
Now for the PC (the "IBM compatibles"): here things are more complicated due to the complexity of the whole platform, i.e. because we have to load an [operating system](operating_system.md) first, of which there can be several, each of which may be loadable from different storages ([harddisk](hdd.md), USB stick, [network](network.md), ...), also we have more complex [CPU](cpu.md) that has to be set in certain operation mode, we have complex peripherals that need complex initializations etcetc. Generally there's a huge [bloated](bloat.md) boot sequence and PCs infamously take longer and longer to start up despite skyrocketing hardware improvements -- that says something about state of technology. Anyway, it usually it works like this:
{ I'm not terribly experienced with this, verify everything. ~drummyfish }
1. Computer is turned on, the CPU starts executing at some initial address (same as with the simple computer).
2. From here CPU jumps to an address at which **stage one [bootloader](bootloader.md)** is located (bootloader is just a program that does the booting and as this is the first one in a line of potentially multiple bootloaders, it's called *stage one*). This address is in the [motherboard](motherboard.md) [ROM](rom.md) and in there typically **[BIOS](bios.md)** (or something similar that may be called e.g. [UEFI](uefi.md), depending on what standard it adheres to) is uploaded, i.e. BIOS is stage one bootloader. BIOS is the first software (we may also call it [firmware](firmware.md)) that gets run, it's uploaded on the motherboard by the manufacturer and isn't supposed to be rewritten by the user, though some based people still rewrite it (ignoring the "read only" label :D), often to replace it with something more [free](free_software.md) (e.g. [libreboot](libreboot.md)). BIOS is the most basic software that serves to make us able to use the computer at the most basic level without having to flash programs externally, i.e. to let us use keyboard and monitor, let us install an operating system from a CD drive etc. (It also offers a basic environment for programs that want to run before the operating system, but that's not important now.) BIOS is generally different on each computer model, it normally allows us to set up what (which device) the computer will try to load next -- for example we may choose to boot from harddisk or USB flash drive or from a CD. There is often some countdown during which if we don't intervene, the BIOS automatically tries to load what's in its current settings. Let's suppose it is set to boot from harddisk.
3. BIOS performs the *power on self test* (POST) -- basically it makes sure everything is OK, that hardware works etc. If it's so, it continues on (otherwise halts).
4. BIOS loads the **[master boot record](mbr.md)** (MBR, the first sector of the device) from harddisk (or from another mass storage device, depending on its settings) into [RAM](ram.md) and executes it, i.e. it passes control to it. This will typically lead to loading the **second stage bootloader**.
5. The code loaded from MBR is limited by size as it has to fit in one HDD [sector](sector.md) (which used to be only 512 bytes for a long time), so this code is here usually just to load the bigger code of the second stage bootloader from somewhere else and then again pass control to it.
5. Now the second stage bootloader starts -- this is a bootloader whose job it is normally to finally load the actual operating system. Unlike BIOS this bootloader may quite easily be reinstalled by the user -- oftentime installing an operating system will also cause installing some kind of second stage bootloader -- example may be **[GRUB](grub.md)** which is typically installed with [GNU](gnu.md)/[Linux](linux.md) systems. This kind of bootloader may offer the user a choice of multiple operating systems, and possibly have other settings. In any case here the OS [kernel](kernel.md) code is loaded and run.
6. Voila, the kernel now starts running and here it's free to do its own initializations and manage everything, i.e. Linux will start the [PID 1](init.md) process, it will mount filesystems, run initial scripts etcetc.
