Simplest Truly Open Computer We Can Make

[tecknian1]

“Computers should be like pencils and hammers: simple tools that anyone can understand, make, repair, and use freely.”

## Ideas for a Public-Domain Computer

Photonic computing could potentially provide very high speeds, low power consumption, and massive parallelism. I would like to discuss whether it could be used to create a practical general-purpose computer, or whether there is a simpler way to achieve the same goal.

This is only an idea for discussion at this stage, not a fixed project plan. I would like to hear people’s thoughts:

- Could a photonic computer be built at home or by a small community?
- Could optical logic be simpler or more powerful than traditional electronic logic?
- What would be the simplest design that could still be fast enough for useful everyday tasks?

The computer should be Turing-complete. I am considering Rule 110, or perhaps something simpler, as a foundation for a design that could simulate a conventional processor using simple building blocks.

It should also be completely modular and flexible, without forcing users to adopt specific types of I/O.

If you are interested in photonics, computer architecture, electronics, software, or radically simpler computers, share your ideas.

[Anon E. Moose]

I have already digged in this direction and my answer, is, unfortunately, "no". photonic computers won't be made from dirt and sticks. though i've seen a guy on yt who described how such computers work and how he is going to build one. however, he already has lithography machine and some other equipment.
for now the simplest way to have open hardware is to build all of the components from transistors and ICs. see pineapple one (homemade 32-bit risc-v cpu) for example: https://pineapple-one.github.io/
i also though about organic semiconductors at some point. for sure, transistors from such components work slower but such transistors manufacture does not require a cleanroom. regarding this part, i found out that copper phtalocyanine is a good semiconductor candidate which is harmless to people, relatively accessible (it's a pigment) and can be synthesized from obtainable components. i can send more links about the process, if anyone is interested. but in order to manufacture transistors the one needs to have a thermal evaporator (expensive but achievable).
there is also a community of people who show how to build tools for silicon chips manufacture (including thermal evaporator): https://docs.hackerfab.org/home

[anonymous_]

Yeah, we need more accessible information about stuff like this. This is the true way of dependency and backdoored-ness.

There's also a surprising numbers of videos on youtube about homebrew CPUs.

[mia2ndpassword]

“Computers should be like pencils and hammers: simple tools that anyone can understand, make, repair, and use freely.”

## Ideas for a Public-Domain Computer

Photonic computing could potentially provide very high speeds, low power consumption, and massive parallelism. I would like to discuss whether it could be used to create a practical general-purpose computer, or whether there is a simpler way to achieve the same goal.

This is only an idea for discussion at this stage, not a fixed project plan. I would like to hear people’s thoughts:

- Could a photonic computer be built at home or by a small community?
- Could optical logic be simpler or more powerful than traditional electronic logic?
- What would be the simplest design that could still be fast enough for useful everyday tasks?

The computer should be Turing-complete. I am considering Rule 110, or perhaps something simpler, as a foundation for a design that could simulate a conventional processor using simple building blocks.

It should also be completely modular and flexible, without forcing users to adopt specific types of I/O.

If you are interested in photonics, computer architecture, electronics, software, or radically simpler computers, share your ideas.

nice concept, but what would you be able to do with them? >:)

[tecknian1]

https://github.com/CMCRobotics/opticalcomputer i thought about that but it is not good enough for home-builders

[tecknian1]

Natural conductors are good for slow tasks. A lot of electronic devices cannot be built at home, such as lasers.

There are also relay computers, but they are slow.

The problem with modern transistors is that they use extremely complex and difficult techniques to make things faster, such as creating nanoscale CPUs and GPUs.

Eventually, things will get worse if no one addresses this crisis.

Air computers exist, but they are not good enough. Optical computers could be affected by dust. What about magnetic computers? magnetic power = electronic power. same issue.

We need a good idea that is both fast and inexpensive.

[tecknian1]

I have already digged in this direction and my answer, is, unfortunately, "no". photonic computers won't be made from dirt and sticks. though i've seen a guy on yt who described how such computers work and how he is going to build one. however, he already has lithography machine and some other equipment.
for now the simplest way to have open hardware is to build all of the components from transistors and ICs. see pineapple one (homemade 32-bit risc-v cpu) for example: https://pineapple-one.github.io/
i also though about organic semiconductors at some point. for sure, transistors from such components work slower but such transistors manufacture does not require a cleanroom. regarding this part, i found out that copper phtalocyanine is a good semiconductor candidate which is harmless to people, relatively accessible (it's a pigment) and can be synthesized from obtainable components. i can send more links about the process, if anyone is interested. but in order to manufacture transistors the one needs to have a thermal evaporator (expensive but achievable).
there is also a community of people who show how to build tools for silicon chips manufacture (including thermal evaporator): https://docs.hackerfab.org/home

can we make a calculator from these materials?

[Anon E. Moose]

Yeah, we need more accessible information about stuff like this. This is the true way of dependency and backdoored-ness.

There's also a surprising numbers of videos on youtube about homebrew CPUs.

Here's what I managed to dig:
https://onlinelibrary.wiley.com/doi/10. ... .202404658 (CF) - the first article found about copper phthalocyanine transistors. A bit scary to read, only last section with "manufacture method" is useful.
https://www.atlantis-press.com/article/125961362.pdf https://pmc.ncbi.nlm.nih.gov/articles/P ... 660952.pdf - other articles about copper phthalocyanine transistors. Used them as a fact-check resources to ensure the article above isn't bs and copper phthalocyanine really can be used for making transistors. Also checked out wikipedia page for this chemical (just in case).
https://docs.hackerfab.org/home/fab-too ... uild-to-do - how to make a thermal evaporator

These videos were used in order to build a chain of reactions for synthesis of copper phtalocyanine:
https://www.youtube.com/watch?v=Mnx6PBFBGik https://www.youtube.com/watch?v=O3IWYr-fv4Q - synthesis of copper phthalocyanine, where are plenty of similar videos.
https://www.youtube.com/watch?v=wQidAZQhxbA - synthesis of phthalic anhydride.
https://www.youtube.com/watch?v=sfldS5hqWYA - synthesis of urea (actually don't bother such experiments, better buy some in shops for gardeners).
https://www.youtube.com/watch?v=T5PJjabZ2EE https://www.youtube.com/watch?v=S4VEZk9ccWg - synthesis of HCl and NaOH, simple but requires some post-preporation steps in order to increase the concentration. Use it only if you aren't able to buy these chemicals directly or obtain from household items (e.g. pipe cleaners).
https://www.youtube.com/watch?v=Ds-jtwXbLAk https://www.youtube.com/watch?v=81jJsDixO70 - more about thermal evaporators.

If I forgot to mention any chemical used in the process, just type "<chemical-name> synthesis" in your favorite search engine or YT. I remember I managed to find all of them.

Also, I'll send here a link when I restore related PDF articles and images from the drive.

Related to efficiency & speed: copper phtalocyanine transistors are much slower than silicon ones. If the second ones are 2-3+GHz, the first ones are only ~20MHz (exact frequency wasn't found anywhere since it wasn't in the focus of articles I found). For simple understanding: you'll have a kind of Arduino in the best case scenario.

Another issue: such transistors are affected by air (not by dust) and slowly degrade with time (much faster than silicon transistors). But if we could create some enclosure (similar to CPUs') this problem would be partitialy solved.

can we make a calculator from these materials?

Yes, if you manage to reliably and repeatedly create transistors.

[tecknian1]

Yeah, we need more accessible information about stuff like this. This is the true way of dependency and backdoored-ness.

There's also a surprising numbers of videos on youtube about homebrew CPUs.

Here's what I managed to dig:
https://onlinelibrary.wiley.com/doi/10. ... .202404658 (CF) - the first article found about copper phthalocyanine transistors. A bit scary to read, only last section with "manufacture method" is useful.
https://www.atlantis-press.com/article/125961362.pdf https://pmc.ncbi.nlm.nih.gov/articles/P ... 660952.pdf - other articles about copper phthalocyanine transistors. Used them as a fact-check resources to ensure the article above isn't bs and copper phthalocyanine really can be used for making transistors. Also checked out wikipedia page for this chemical (just in case).
https://docs.hackerfab.org/home/fab-too ... uild-to-do - how to make a thermal evaporator

These videos were used in order to build a chain of reactions for synthesis of copper phtalocyanine:
https://www.youtube.com/watch?v=Mnx6PBFBGik https://www.youtube.com/watch?v=O3IWYr-fv4Q - synthesis of copper phthalocyanine, where are plenty of similar videos.
https://www.youtube.com/watch?v=wQidAZQhxbA - synthesis of phthalic anhydride.
https://www.youtube.com/watch?v=sfldS5hqWYA - synthesis of urea (actually don't bother such experiments, better buy some in shops for gardeners).
https://www.youtube.com/watch?v=T5PJjabZ2EE https://www.youtube.com/watch?v=S4VEZk9ccWg - synthesis of HCl and NaOH, simple but requires some post-preporation steps in order to increase the concentration. Use it only if you aren't able to buy these chemicals directly or obtain from household items (e.g. pipe cleaners).
https://www.youtube.com/watch?v=Ds-jtwXbLAk https://www.youtube.com/watch?v=81jJsDixO70 - more about thermal evaporators.

If I forgot to mention any chemical used in the process, just type "<chemical-name> synthesis" in your favorite search engine or YT. I remember I managed to find all of them.

Also, I'll send here a link when I restore related PDF articles and images from the drive.

Related to efficiency & speed: copper phtalocyanine transistors are much slower than silicon ones. If the second ones are 2-3+GHz, the first ones are only ~20MHz (exact frequency wasn't found anywhere since it wasn't in the focus of articles I found). For simple understanding: you'll have a kind of Arduino in the best case scenario.

Another issue: such transistors are affected by air (not by dust) and slowly degrade with time (much faster than silicon transistors). But if we could create some enclosure (similar to CPUs') this problem would be partitialy solved.

can we make a calculator from these materials?

Yes, if you manage to reliably and repeatedly create transistors.

it will be good if we can make a computer without plastic tools.

also this speed isn't really perfect, it can be good for calculators and small tasks but not for something like general-purpose computer.
Arduino speed will be decent for a lot of things that we do daily like reading articles.

how big this transistor will be? millimeter scale?

[Anon E. Moose]

it will be good if we can make a computer without plastic tools.

Don't quite understand what you mean.

how big this transistor will be? millimeter scale?

Start with making at least cm scale, to test things out. It should work at 40V or so.
Here's the link for all the documents regarding this type of transistors I found: https://upload.disroot.org/r/AJ4qNDWv#b ... YtJp8nwJI=
Also, if you want to do some practical stuff, it's better to begin with buying already synthesised CuPc. It's much cleaner and cheaper than the synthesis option if you live somewhere where chemicals are relatively easy accessible. Unfortunately, I couldn't test manifacture of such transistors because chem. vendors in my country don't sell even such harmless stuff.

[tecknian1]

It would be good if we could make a computer without using plastic tools.

I don't quite understand what you mean.

I meant that I would prefer a fabrication process that does not involve unpleasant or hazardous chemical fumes.

How large would this transistor be? Would it be on the millimeter scale?

Start by making it at least centimeter-scale so you can test things out. It should work at around 40 V.

Here's a link to all the documents I found regarding this type of transistor: https://upload.disroot.org/r/AJ4qNDWv#b ... YtJp8nwJI=

Also, if you want to do something practical, it's better to start by buying already-synthesized CuPc. It's much cleaner and cheaper than synthesizing it yourself if you live somewhere where chemicals are relatively easy to access. Unfortunately, I couldn't test the manufacture of such transistors because chemical vendors in my country don't even sell such harmless materials.
[/quote]

I would need to research this more.

I mainly wanted to build a computer with as little electronics as possible, so I thought about optical computers or optomechanical computers, They use light rather than electronics for computation, which means they can be more power-efficient and faster, but they are difficult or expensive to build.

[tecknian1]

I think the key distinction is between a **purely optical computer** and a **home-buildable computer that uses optical signaling**.

A pure photonic CPU is probably not the simplest route. Optical logic requires thresholding, gain, fan-out, memory, and signal regeneration. Lenses and mirrors alone do not solve these problems. Dust is less important than alignment, signal loss, and the lack of practical optical memory.

A more realistic open design would be a small **Minsky two-counter machine** built from ordinary logic, LEDs, photodiodes, and ROM. It would have:

- two expandable binary counters;
- increment and decrement operations;
- zero detection;
- conditional branching;
- a small programmable instruction memory;
- modular I/O;
- optional optical links between modules.

A first version could use two 8-bit counters, a small ROM program store and a manually stepped clock. It would implement the basic two-counter-machine model, although any physical version would have finite storage. The counters could later be expanded to 16 or 32 bits.

The optical part could be added without requiring lithography. This would make the machine optoelectronic rather than purely photonic, but it would remain cheap, modular, repairable, and buildable by a small community. LEDs, photodiodes, cardboard optical tubes, breadboards are much more accessible than lasers, custom waveguides, nonlinear optical materials, or a home semiconductor fab.

I would divide the project into stages:

1. Build a 4-bit two-counter machine.
2. Add a programmable ROM instruction store.
3. Expand the counters to 8 or 16 bits.
4. Add a modular I/O bus.
5. Replace selected electrical links with LED/photodiode links.
6. Experiment with more optical logic only after the electronic reference design works.

This also gives us a useful answer to “what would people do with it?” It could run interpreters, calculators, simple games, control systems, and educational software. It would not compete with a modern CPU, but it could be a genuinely open computer whose architecture, firmware, construction, and repair process are all understandable.

[tecknian1]

give the link of the yt guy

I have already digged in this direction and my answer, is, unfortunately, "no". photonic computers won't be made from dirt and sticks. though i've seen a guy on yt who described how such computers work and how he is going to build one. however, he already has lithography machine and some other equipment.
for now the simplest way to have open hardware is to build all of the components from transistors and ICs. see pineapple one (homemade 32-bit risc-v cpu) for example: https://pineapple-one.github.io/
i also though about organic semiconductors at some point. for sure, transistors from such components work slower but such transistors manufacture does not require a cleanroom. regarding this part, i found out that copper phtalocyanine is a good semiconductor candidate which is harmless to people, relatively accessible (it's a pigment) and can be synthesized from obtainable components. i can send more links about the process, if anyone is interested. but in order to manufacture transistors the one needs to have a thermal evaporator (expensive but achievable).
there is also a community of people who show how to build tools for silicon chips manufacture (including thermal evaporator): https://docs.hackerfab.org/home