I had used julia v0.6 to write some physical modeling code. The code wasn’t used for quite some years (my PhD was over). We recently decided to dust it off and upgrade it to work with julia v1.10, so it would be easy for more people to install and use etc.There really wasn’t so much to change (a few imports, linspace to range, …).
Installation ‘just works’. The results are identical to within machine precision. It runs between x2 and x10 faster now.
Wow, that must feel like using almost an entirely new language! I played around with Julia at v0.7 for a bit, then didn’t revisit until v1.4, which already felt like a massive improvement.
I was using the following packages: DSP, IterTools (not sure why), ProgressMeter and YAML. We added: FFTW, DelimitedFiles, Revise, Rsvg and Runner. A few imports had to be added from the standard library.
A few demonstration scripts use PlotlyJS and Plots, but I don’t think I tested those. However, it was rather simple use, so no worries there.