seshat

This software provides a general-purpose data and software repository system that attempts to promote active re-use of data and software by providing a rock-solid foundation of implementations of protocols and standards related to its aims.

Reporting (potential) security issues

For security-related matters, please e-mail security@seshat.software. If you are (going to be) running an instance of this project and would like to be part of the security mailing list, then please write the mailing list.

Production-ready containers and packages

Container-based

Pull the latest image using:

docker pull seshat.software/seshat:latest

Package managers

RPM packages for Enterprise Linux and Fedora can be obtained from COPR.

Debian packages can be obtained from the Codeberg releases page.

PyPI

Installing via PyPI can be done using:

pip install seshat

Creating a development environment

This project uses the GNU autotools build system.

GNU/Linux

For development on GNU/Linux we recommend installing git, autoconf, automake, make and typst through your system's package manager, followed by creating a Python virtual environment for seshat:

git clone https://codeberg.org/seshat/seshat.git && cd seshat/
autoreconf -if && ./configure
python -m venv ../seshat-env
. ../seshat-env/bin/activate
pip install --upgrade pip
pip install --editable .

macOS X

For development on Apple's macOS X, we recommend installing python3, git, autoconf, automake, and make through homebrew, followed by creating a Python virtual environment for seshat:

brew install python3 git autoconf automake make typst
git clone https://codeberg.org/seshat/seshat.git && cd seshat/
autoreconf -if && ./configure
python3 -m venv ../seshat-env
. ../seshat-env/bin/activate
pip install --upgrade pip
pip install --editable .

Microsoft Windows

For development on Windows we recommend MSYS2 and the following approach to installing packages:

PRE="mingw-w64-ucrt-x86_64-" # See https://www.msys2.org/docs/package-naming
pacman -Suy auto{conf,make} make ${PRE}{git,typst,python,python-setuptools} \
       ${PRE}python-{defusedxml,jinja,pillow,pygit2,rdflib,requests,werkzeug,pip}
# If you chose a different PRE above, change /ucrt64 accordingly below.
# See: https://www.msys2.org/docs/environments
export PATH="/ucrt64/bin":"${PATH}"
git clone https://codeberg.org/seshat/seshat.git && cd seshat/
python -m venv --system-site-packages ../seshat-env
. ../seshat-env/bin/activate
autoreconf -if && ./configure
pip install --editable .

Tip: On Windows, prefer 127.0.0.1 over localhost in the sparql-uri and sparql-update-uri configuration. Windows resolves localhost to IPv6 (::1) first, and when the SPARQL endpoint only listens on IPv4, every request waits for the failed IPv6 connection attempt before falling back, adding a second or more to each query.

Keeping your development environment up-to-date

GNU/Linux and macOS X

Because the virtual environment isn't updated by your system's package manager, you can use the following snippet to update packages inside your virtual environment:

pip install -U $(pip freeze | grep -v "seshat.git" | cut -d= -f1)

Microsoft Windows

The dependencies for seshat are installed via pacman, so to update those packages use the following snippet:

pacman -Suy

See Updating MSYS2 for more details.

Verify that the installation works

Upon completing the installation, you should be able to run:

seshat --help

Setting up the database

Seshat needs a SPARQL 1.1 endpoint such as Virtuoso OSE or Jena Fuseki to store its state.

Run the web service

To start the web service, we recommend copying the example configuration and go from there:

cp etc/seshat.xml seshat.xml

First run

Upon first run, seshat needs to initialize the database with categories, licences and accounts. To do so, pass the --initialize option to the seshat web command:

seshat web --initialize --config-file seshat.xml

Subsequent runs

After the database has been initialized, you can remove the --initialize option:

seshat web --config-file=seshat.xml