That sounds like a great idea! To prepare for that, I’ve pushed a commit to include a “date” column with the release date (in format yyyy-mm-dd, i.e. ISO 8601) in the generated .csv file for each benchmarked version; having that it should be relatively easy to make such a graph. I haven’t adapted plot.py yet to make use of that new column, will look into that later that week. If you or anyone else is eager to give it a shot already (my plotting skills are close to zero, the code is largely derived from ChatGPT interaction), feel free ![]()
2 Likes