Civboot: a civilizational bootstrapper
What I cannot create, I do not understand - on the blackboard of Richard P. Feynman at the time of his death.
Everything should be made as simple as possible, but no simpler. -- Albert Einstein
This is the repository for Civboot: a Civilizational Bootstrapper. It's purpose is to act as a version-controlled wiki for all efforts related to Civboot.
This effort was started on 2019-11-14 and is still very early stage. This github repo should be used for discussion (through opening an issue).
I have also started writing fiction.
How large would a warehouse have to be in order to both:
- Build simple computers in a semi-autonomous way from scratch using only tools and knowledge within the warehouse.
- Build another such warehouse and make improvements to it.
If such a warehouse existed, it would be a Civboot. Civboot aims to simplify technology by simplifying technology's requirements. Fundamentally Civboot is an educational tool: it can provide a working understanding of the tools and processes that underpin our technology. Putting the question another way, if there were a high-school & college curriculum with the sole goal of accomploshing "project based learning" of having the entire class (400+ students) understand and reconstruct a Civboot using only the tools within the Civboot; how many people would that be, how large would the warehouse be and how many instructors would be required?
This is the fundamental question of Civboot, a project which aims to reduce the tools and knowledge necessary to bootstrap modern civilization as much as possible. It is important to point out early that a Civboot does not contain everything required to build or support Civilization such as food, clothing or a bed; in the same way as a bootstrapper for an operating system does not contain all the bells and whistles of the operating system itself. A Civboot only contains the tools and knowledge necessary to build computers and itself.
In terms of supporting a community with things like food, clothing and shelter, a Civboot has one additional goal: be able to provide these things within three degrees. For instance, a Civboot cannot construct a tractor directly, but you can use tooling within a Civboot to build a factory (the 1st degree), then you can use the factory to build gears and other tractor components and assemble a tractor (the 2nd degree).
Another point: a Civboot only contains tutorials to provide a working knowledge of how to reconstruct and improve the Civboot. Deeper understanding of things like transistors, hardware architecture, software architecture, chemistry, mechanical engineering, etc can take potentially a lifetime of learning. It is hoped that Civboot can be a foundation for that learning, but Civboot itself aims to provide only a working understanding.
This project is NOT an attempt to "do this at home." Creating the first Civboot will no doubt require millions of dollars and the efforts of thousands of individuals to design, experiment, document and simplify a huge number of technologies which cross an extremely broad range of disciplines. The first Civboot will likely be made "distributed" (if it is ever made at all) -- with teams from all over the world sending their products to each other, collaborating on multiple different approaches. The hope is to get hackers, professionals, retired experts, universities and organizational support to help distill civilization's knowledge into a stack that can be understood and self replicated.
Even if we fail, it will be worth the effort. Having a simplified working-understanding of technology has many advantages other than the construction of a Civboot.
Philosphy of Civboot
Civboot is not a criticism of the current complexity of technology. Technology is (largely) as complex as we want it to be for the requirements we have: hiding internal complexity, simplified and beautiful user interfaces, and scalability of more engineers working on large-scale problems.
Civboot is founded on a philosophical idea: modern civilization would be better off if it owned the means to produce it's essential components. This is not to say that proprietary pieces of the tech stack are somehow wrong (although some believe that). Civboot is not an economic philosophy, it is a scientific one. Humanity is better off when it can understand the technology it depends on, and we can only understand what we can create.
You are probably thinking the real goal of Civboot is to be able to survive an apocalypse-type scenario. This was indeed one of the inspirations for Civboot, and the project CollapseOS was a motivation for the current technology stack that Civboot is pursuing. I believe that thinking along "negative" outcomes such as "what happens if civilization collapses" is highly productive to creativity, growth and learning in much the same way as death and rot are essential for the growth of plants and sustenence of all life. Having a noble goal of being educational and a motivational goal of concern for the future provides a good balance from which I hope Civboot and its community can thrive.
Civboot is and must remain completely free and open. In order to fulfill its mission, it must be possible to have a working-understanding of every level of its design -- from materials science, to manufacturing process, to mechanics, to the computer hardware and software it is running. Included information and processes should be as simple as possible -- these concepts should be easy to learn and replicate. This constraint will force simplification of the designs, ideas and software being included. We must take our complicated technology stack and work on distilling it to its essentials, then rebuild upon those essentials to be even more efficient and simple.
The first goal of Civboot is to be able to create a technology stack which can self-replicate with 500 motivated students in 8 years, starting with a US-equivalent 8th grade education. In addition, the goal is to be able to rebuild a Civboot (with a Civboot) for less than 25 million dollars -- or $50,000 per student, less than the typical cost of 8 years of education in the United States.
In addition, there are (aspirationally) 5 Civboot levels. These are not part of the core spec, but may help get creative juices flowing. They are documented here
The intent of these constraints is to limit the amount of tools and information included in a Civboot into a manageable amount. These constraints (whether included or excluded) make no judgement about the importance of said items except as to apply to the goal of creating the first Civboot.
- Must be entirely free and open.
- Must not contain proprietary or hidden designs, i.e. no "binary blobs"
- Must be possible to build itself from itself
- Requires no outside knowledge besides the language which the guides are written in
- Requires no outside tooling besides commonly found materials (rocks, clay, wood, dirt, sand, ore, etc)
- Must include all working knowledge necessary to understand every
component and their interactions.
- Must include the theory and practice of math, physics, chemistry, materials science, computer design and programming, etc necessary to explain the design of every component in the system such that it can be replicated and have improvements made upon it.
- Must not include information not necessary to understand or begin
improvements upon the system. Aka must not include politics, history,
medicine, philosophy, religion, art, psychology, sociology or any branch of
human knowledge or endeavor outside of the sciences and engineering
necessary to understand and improve the system. These are important to
preserve, but it is not the goal of Civboot to preserve them.
- Civboot should aid in providing the technology to store and view such knowledge where possible.
- May include some amount of history of math/science/engineering/philosophy which is required to properly understand a component.
- Should be as simple as possible but no simpler.
- Comfort has almost no importance, unless it hinders use excessively
- Easy as possible to re-learn and re-create from near-scratch
- Should be as efficient as possible
- Require as few materials / space / energy necessary to achieve its purpose. Improvements to these areas without increasing complexity should be accepted.
- Should focus on the "single city" design
- Local communication protocols acceptable
- Globalized transportation / communication not required (but allowed)
- Should be composed of modifiable and swappable designs
- Stack will continue to improve in both simplicity and effectiveness -- no stage in stack should block this improvement
- Stack should be upgradeable as components become simpler and more efficient
- May include multiple (swappable) alternatives to perform same role with different tradeoffs
- Should not include knowledge or tools just for the sake of those who
- For example, although the people working on a new Civboot need to eat, farming equipment and practices should not be included.
- Some equipment, like a bulldozer/truck, have multiple purposes but may be required to build the Civboot itself so should be included.
- Civboot should enable the storage/retrieval of such knowledge and construction of such tools
The following is an effort to compile what is needed to construct a Level 1 Civboot and the current status.
General Tools and Equipment (non computer): See Open Source Ecology for current work on many of these designs. These two movements share a lot in common.
- Human Tools and Equipment:
- Hand tools: saw, hammer, axe, shovel, pickaxe, etc
- Powered hand tools: drill, rotary saw, soldering iron, etc
- Vehicles: tractor, bulldozer, etc
- Materials creation: blast furnace, forge, brick builder, plastic (like) material synthesis, etc
- Fabrication: CNC machine, 3D printer (multi-material), lathe, etc
- Power Usage: electric motor (+generator)
- Power Generation: Steam generator, wind, heat exchanger (sterling engine)
- Power Storage: battery, steam storage, gravity storage
Computer Hardware: the computer hardware manufactured will likely be extremely limited compared to modern computer hardware.
- podcast 0002 focused on whether semiconductors can be built.
- Basic requirements are to either simplify the processing of Silicone or a different semiconductor material (like Tin(II) Sulfide) and/or find a simpler process such as nanotube CPU. There is some hope in Thin Film Transistors which can be made on glass (NOT pure silicone) using low-temperature processes.
- This is the hardest piece of a civboot.
- Once circuits can be constructed, a simple stack-based CPU like the J1 or the scamp-cpu is likely the best target for Civboot, as such CPUs can be produced with only a few thousand transistors.
Computer Software: CollapseOS gives a pretty good idea of how to get software bootstrapped on basic hardware using only a few hundred lines of assembly and forth code. That OS fits on less than 5k, requires no binary blobs, and can cross compile to several other architectures. The Forth programming language is a primary driver of this simplicity. A (forth) compiler, basic file system, text editor, assembler and cross compiler was written largely by a single individual and runs on a computer chip with <9000 transistors from the 1970's.
The basic language stack for software is currently imagined to be: Forth OS / bootloader like CollapseOS. This will be used to build a new computer language fngi, which can be used to make the full OS and coding environment.
On top of this, software should be written including CAD, board-layout and VLSI software similar to ImplicitCAD, i.e. does not require a GUI. In addition communication protocols will likely be similar (or identical) to the PJON 1-wire protocol and Ham Radio, with minimal support for TCP/IP. Finally, a site-database can hold tutorials and educational software similar to Khan Academy, which can aid the self-learning of Civboot students.