Friday, 29 August 2025

Just a blorb to remind me of the presentation by Brian Goetz during 2025 JVM Language Summit.

which was made available on the YouTube¹.

P.S. in the link above was a reference to a presentation (in written form) by Guy Steel², that was insightful.

References

: https://youtu.be/Gz7Or9C0TpM?si=ejyjWLRKwSY05AKW
[2] Growing a Language - Guy L. Steel Jr.: https://www.cs.virginia.edu/~evans/cs655/readings/steele.pdf

Occasionally, I have to scan official papers (dead-tree-copy) and I'd like to keep the result small, and in one pdf.

So, a one liner command would be nice.

magick mrbear*.png -quality 10% -set filename:base "%[basename]" "%[filename:base].jpg"

The commandline above automatically creates an appropriate .jpg file for each mrbear*.png file. So mrbear_0001.png will cause a mrbear_0001.jpg.

Just what I wanted. Went from 8 megabytes picture to a 200 kb picture.

And then converting the new images over to a pdf file is ridiculously simple:

magick mrbear*.jpg mrbear.pdf

Magick!

DigitalOcean - Reduce File Size of Images in Linux - CLI and GUI methods: https://www.digitalocean.com/community/tutorials/reduce-file-size-of-images-linux
GitHub - Convert images but keep the original file name #6494: https://github.com/ImageMagick/ImageMagick/discussions/6494
ImageMagick: https://imagemagick.org/