April 16, 2026

Snakes, labs, and comment wars

Python Package Compiler:Package Matlab Programs for Deployment as Python Package

Matlab crashes the Python party — and the comments shout “just use Julia”

TLDR: Matlab launched a new app to package Matlab code as Python modules in R2025a, aiming to make old projects play nice with modern Python apps. Commenters immediately clashed: practical bridge for legacy code vs. “just rewrite it in Julia,” with side debates on speed, licensing, and deployment hassle.

Matlab just dropped a new trick: a point‑and‑click “Python Package Compiler” that lets you bundle Matlab code and ship it as a Python package. In simple terms, it’s a bridge so teams with years of Matlab work can plug straight into the massive Python world. It arrives in the R2025a release with a fresh app, project saving, and a tidier output folder system, replacing the old Library Compiler for this job. Official word is all menus, one interface, less hassle — you click, it builds, you share MathWorks.

But the internet heard “Matlab” and instantly split into camps. The loudest early take? “Skip the bridge, just rewrite it in Julia.” One veteran commenter boasted they ditched Matlab 12 years ago and never looked back, claiming wild speed gains and the same hackable feel. Cue eye‑rolls from folks who live with 15‑year‑old research models and toolbox dependencies: “Rewrite in 20 minutes? Sure, for a homework problem.” Others cheered the move as a lifesaver for mixed teams: let scientists stay in Matlab while Python apps get the goods.

Memes flew fast: “Google Translate for code,” “Matlab cosplay as a Python package,” and inevitable licensing jokes. Skeptics flagged the runtime weight and deployment friction; supporters countered that installers and persistent projects are exactly what big orgs need. Verdict: the feature is practical — the culture war is the show.

Key Points

  • Python Package Compiler is a MATLAB app introduced in R2025a to package MATLAB programs as Python packages.
  • The app’s UI builds customized compiler.build commands via interactive menus and dialog boxes.
  • Advantages include a single deployment interface, project-based organization, persistent project state, task reuse, and installer packaging.
  • The app can be opened via the MATLAB toolstrip (Apps > Application Deployment) or by entering pythonPackageCompiler at the MATLAB prompt.
  • From R2025a, output files are generated in a folder named after the compiler task, replacing earlier folder names used up to R2024b.

Hottest takes

“Or spend 20 minutes to rewrite your Matlab as a Julia program” — slwvx
Made with <3 by @siedrix and @shesho from CDMX. Powered by Forge&Hive.