This week has been about the SQL DIFF operator, Nix flakes, JupyterLab spreadsheets, browser privacy tests, graph-based ASTs, a reflection on Markdown.
# Thoughts On Markdown
An article reflecting on almost 20 years of Markdown, whether it's original intent is (still) valid and what could be a way forward with a balanced set of features, extensibility and ease of entrance.
I found the article initially thought-provoking but rapidly deriving into a niche solution for solving the author's employer problems. It completely ignores situations like offline-first, overemphasises the need for real-time collaborative experiences and ignores XML all together which is at the very least a precursor of Portable Text, the solution he is pitching for.
# Nix Flakes: an introduction
An article on Nix flakes covering its objective, basic setup to enable the feature whilst it's behind the experimental flag and a run-through of using them with the Go language.
A JupyterLab plugin to edit Pandas dataframes as spreadsheets and generate Python code corresponding to each edit.
# Type inference in ogma using graphs
An article exploring graph-based ASTs for partial compilation-based type inference. ogma, a scripting language for processing tabular data, is worth a look as well.
# Data diffs: Algorithms for explaining what changed in a dataset
An article introducing the DIFF operator for relational databases to analyse changes in schema-aligned datasets.
# Privacy Tests
An open-source initiative that subjects popular web browsers to a suite of automated tests designed to audit web browsers' privacy properties.