Evaluate Migrating from Webpack (via django-webpack-loader) to Vite in our Django stack

[pyanderson]

I’d like to propose evaluating a migration (or partial adoption) from our current Webpack + django-webpack-loader setup to Vite, in order to improve developer experience, build performance, and modern tooling. However, given our Django integration, the migration will require special care and planning.

📌 Context & Constraints (Django + django-webpack-loader)

• We currently use django-webpack-loader to bridge Django templates with Webpack-built assets, by reading a stats/manifest file and injecting script/css tags at runtime.
• This coupling means that any replacement must still allow Django to reference built assets (with hashed filenames, splits, etc.).
• Some Webpack-specific plugins, loaders, or configurations (e.g. custom module resolution, asset handling) may not have direct counterparts in Vite.
• We also often rely on features such as hot reload / HMR in dev mode, and stable builds in production that mesh with Django’s static file system and collectstatic.
• So migrating is not a "drop-in swap"; we’ll need to address how Django will consume Vite assets, replace or replicate webpack-loader behavior, and adapt existing pipeline scripts.

✅ Potential Benefits / Pros

1. Faster development builds & better hot-reload
   • Vite’s dev server with native ES modules can reduce cold-start times and speed up incremental reloads.
   • HMR tends to be more responsive in Vite setups.
2. Simpler configuration & less boilerplate
   • Vite often requires less configuration for common patterns, reducing long Webpack configs.
   • Many modern features (e.g. code splitting, tree shaking) are built in.
3. More alignment with modern frontend tools
   • Easier compatibility with ESM-based libraries and tools.
   • Growing ecosystem around Vite / Rollup / plugins.