What codebase would you like to see rewritten, updated, or modernized?

Question

Inspired by this phenomenal write-up of the author's experience in rewriting tmux[1], I'd like to hear from HN: what codebases would benefit from a similar treatment? Or what codbases would you like to see get the 'star' treatment in some way or another -- an upgraded tech stack, modernization, or a rewrite in another language, or in a different idiom/pattern, that you think would be better suited to the job it does?[1] https://news.ycombinator.com/item?id=44455787

colkassad · Accepted Answer

GDAL [1]Don't get me wrong, it's a fantastic and ancient coral reef of a library and tool set but has a lot of inconsistencies in naming conventions and usage (gdal_translate, gdalfinfo). Many geospatial professionals are not savvy enough to leverage it without it being wrapped by someone else like Esri. Windows power users typically install it with something like OSGeo4W [2], whose name I can never remember. Whenever I need it I spin up a Docker image for convenience.[1] https://github.com/OSGeo/gdal [2] https://trac.osgeo.org/osgeo4w/

horsellama · Answer

homebrew. I dream of something quick and simple to use as uv is now for python ecosystem

geophph · Answer

The main codebase a team I work with uses.They&rsquo;re a data science team mostly but the main &ldquo;data processing engine&rdquo; that they solely rely on was written by a dude who split right as he &ldquo;finished it&rdquo;. No one knows how it really works. Black box style. Has been &ldquo;the way&rdquo; for 5 years. I&rsquo;m not sure if no one cares, or no one cares to care, but they don&rsquo;t seem to want to change it. Maybe it works, but also funny how many business critical decisions are made based on the code that everyone has only ever &ldquo;just trusted&rdquo;.