Skip to content

Latest commit

 

History

History
128 lines (91 loc) · 10.3 KB

README.md

File metadata and controls

128 lines (91 loc) · 10.3 KB

Overview

Psibase is a powerful protocol that enables communities to come together to easily self-host web applications. This repository is an open source implementation of that protocol. Any psinode deployment comes with some default applications, some of which are mandatory and critical for the infrastructure to function properly, and others are optional sample user applications.

Primary packages included in this repository:

Package Description
psinode The psibase infrastructure node
psibase Command-line utility for interacting with psinode. Used primarily by scripts.
psidk The SDK used for developing web services

Demo deployment 🖥️

A demonstration deployment is currently hosted at psibase.io to showcase what psibase infrastructure could look like for end users. The applications available on this demo are sample applications and are not intended to be used for a production deployment.

Technical docs 📓

Documentation is hosted on the docs app of our demo deployment.

Build from source 🛠️

We only officially support developing natively on Ubuntu 20.04 and Ubuntu 22.04, although any Linux distribution may work. If you use a distribution other than Ubuntu 20.04 or 22.04, ensure the versions of the dependencies you install are at least as high as they are in Ubuntu.

Some of our developers have very strong success with Ubuntu on WSL2, and also with Docker.

Environment setup

📦 We maintain containerized development environments to simplify the development process. Developing within a container means the environment will automatically be set up for you. See the Psibase Contributor repository for instructions on how to build from source within a container.

The following is a list of dependencies you are required to install:

Package Minimum Version Description
autoconf 2.69 A tool for producing configure scripts for building, installing, and packaging software
binaryen 113 A compiler and toolchain infrastructure library for WebAssembly
boost 1.78 A collection of C++ libraries
clang + llvm 16 The LLVM compiler and toolchain for C/C++ and the libraries to interface with it
cmake 3.16.3 A build system generator that creates makefiles used for compilation
g++ 11 A GNU compiler for C++
libssl-dev 1.1.1 Development package for OpenSSL, enabling various cryptography and secure network communication
libtool 2.4.6 Script that simplifies the process of creating and using shared libraries
make 4.3 Reads makefiles generated by cmake to build executables from source code
node 16.17.0 A JavaScript runtime
pkg-config 0.29.1 A helper tool used during compilation to provide correct compiler flags, and to locate and manage library dependencies
Rust toolchain 1.78 Develop programs in Rust. Use cargo to install cargo-component, mdbook, mdbook-linkcheck, mdbook-mermaid, mdbook-pagetoc, sccache, and wasm-pack
Wasi SDK 20.0 A development toolchain for building WebAssembly programs that target the WebAssembly System Interface (WASI)
zstd 1.4.4 Command-line utility that implements the Zstandard file compression algorithm

Other notes:

  • Add the paths to node, npm, yarn, cargo, and wasm-opt into your PATH environment variable.
  • Set the WASI_SDK_PREFIX environment variable before building to the root of your llvm installation (e.g. /usr/lib/llvm-16). Alternatively, set this variable during the build using the CMake flag -DWASI_SDK_PREFIX=.....
  • Add the following targets to rust: wasm32-wasi, wasm32-unknown-unknown

🔍 You can reference either the Ubuntu 20.04 dockerfile or the Ubuntu 22.04 dockerfile for an example on how an environment could be set up.

Build

After your environment is set up, then you should be able to use the following build instructions:

git submodule update --init --recursive
mkdir build && cd build
cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=psidk ..
make -j $(nproc)
ctest -j $(nproc)
make install

The built product lives in CMAKE_INSTALL_PREFIX ($CMAKE_BINARY_DIR/psidk in the above example).

Build flags

The following are some of the common/useful build flags you can use to configure CMake that modify the build. Open the build directory and run cmake -LH to view a list of all options.

Flag Example Values Description
BUILD_DEBUG_WASM yes,no Build debug wasms (Needed for using the debugger)
CMAKE_BUILD_TYPE Debug,Release Determines level of optimization for the native programs
CMAKE_EXPORT_COMPILE_COMMANDS yes,no Tell CMake to generate compile_commands.json, used by various tools (e.g. clang format intellisense)
CMAKE_INSTALL_PREFIX psidk Binary install location.

Running psibase infrastructure 🚀

🔍 We currently do not provide an ARM binary or docker image. If your device uses an ARM chip, you should build psinode (or its docker image) yourself.

To run psinode, you can either build from source as described above, or you can install prebuilt binaries through one of the following two methods:

  1. Installing a binary release, the latest of which can be found on our Releases page. We provide binary packages for both Ubuntu 20.04 and Ubuntu 22.04.
  2. Installing our psinode Docker Image.

To execute psinode from the docker image, you can use the following docker command (make sure to specify the VERSION you want to run):

docker run --rm -p 8080:8080 -p 3000:3000 -p 9090:9090 ghcr.io/gofractally/psinode:VERSION

Run psinode with --help to learn about the various options, or consult our technical docs.

Contribute 🧬

If you're interested in developing next-generation web application development & deployment infrastructure, welcome aboard! Come hang out in the community telegram channel to learn how best to contribute.

Alternatively, fork the project and do whatever you want with it!

MacOS 🍏

If you have a Mac (including Apple Silicon), you can use an Ubuntu 22.04 VM or Docker container. We do not host ARM variants of our docker images on our GitHub container registry. Therefore if you would like to use either the psibase-builder-ubuntu-2204 or psibase-contributor on your Apple Silicon Mac, you should pull the image-builders repo and build the images yourself.

Support 📱

Some contributors to this repo hang out in this telegram channel. That's probably the best place to ask questions and get support.