I ran an experiment yesterday, and found the package with the largest number of dependencies in #guix using this little command: guix package -A | awk '{print $1}' | parallel 'echo -n {}; echo -n " "; guix size {} 2>/dev/null | tail -n +3 | wc -l' | sort -n -k 2 | tee ~/Documents/sorted_packages.txt. The whole thing took 6 hours to run at about 90% CPU utilisation.
Result: pigx has the most dependencies at 968, with most of the runner ups being scientific analysis tools too. Cool!
I love how #guix allows you to inspect all of the dependencies of a package in detail. Running guix graph foo | xdot - on some package and combing through the massive dependency graph that results is fascinating, but also a little terrifying.
It's a wonder anything works at all.
Add comment