Installation

Dependencies

  • Kernel: Linux 3.9 or newer.

  • The psmisc utilities.

  • The development headers for SQLite, version 3.7.0 or newer.

  • Python 3.10 or newer, installed with development headers.

  • Additional Python dependencies are listed in pyproject.toml (and your Python package installer should normally be able to install them automatically)

Installing from a source release

Download the release tarball s3ql-A.B.C.tar.gz and its detached signature s3ql-A.B.C.tar.gz.sig from GitHub and validate them with signify:

signify -V -m s3ql-A.B.C.tar.gz -p <public-key>

The public key must be obtained from a trustworthy source for the first installation. After that, the trust chain is self-sustaining: each release tarball’s signify/ directory contains public keys for future releases. Bugfix releases (A.B.C) are signed with s3ql-A.B.pub; new major or minor releases (A.B.0) are signed with s3ql-next.pub. After verifying a tarball, extract the relevant key from its signify/ directory and keep it for the next verification.

After validating the tarball, unpack it and change into the newly created s3ql-X.Y.Z directory. At this point, you can use any tool that can work with Python pyproject.toml files to install S3QL. If you don’t know which one to use, pipx is a good choice. To install S3QL with pipx, run

pipx install .

This will download all necessary Python dependencies and install S3QL in its own separate virtual environment.

For a minor performance boost, set the PYTHONOPTIMIZE environment variable before running S3QL commands (perhaps by using a wrapper script). This disables the checking of debug assertions. Surprisingly, there does not seem any way to do this automatically for a Python application (otherwise S3QL would set this up by default).

Installing from Git / Developing S3QL

Clone the S3QL repository and take a look at developer_notes/setup.md.