Why Plasmo uses Parcel instead of Vite for bundling?

[only-issues]

I've been working with Plasmo for a while and really enjoy the framework. However, I noticed that the development experience is somewhat slow due to Parcel's rebuild times. I've also tried WXT (which uses Vite), and the rebuild times are significantly faster.

Current observation:

• Plasmo uses Parcel as its bundler
• Development rebuild times with Parcel are noticeably slow
• Comparative experience with WXT (Vite-based) shows much faster rebuild times

Questions:

1. What were the technical considerations behind choosing Parcel over Vite?
2. Are there any specific advantages that Parcel provides for browser extension development that Vite doesn't?
3. Is there any possibility of having Vite as an alternative bundler option in future releases?

I understand there might be technical limitations or specific requirements that led to this choice, but I'm curious about the reasoning and if there are plans to explore faster bundling alternatives in the future.

[oyzamil]

In my opinion, the Plasmo team chose Parcel mainly because it’s zero-config and handles many browser-extension edge cases automatically (multiple entry points, background/content scripts, HTML pages, MV2/MV3 differences) without requiring custom config.

Parcel also has strong built-in support for:

• Multi-bundle dependency graphs
• Web extension targets
• Automatic asset handling (CSS, static files, workers)
• Minimal configuration for beginners