Hacker News new | past | comments | ask | show | jobs | submit login

Per comments above, pypy.js is not presented as an trivial but interesting thing, but as something useful.



It's easy to see how this could be useful for somebody though.

Imagine you're a scientist, and you have a large, pure python codebase - let's say for intelligently predicting the number of foos in a bar - and you want to make that available via the browser, in order to visualize the progress of your foo-hunter.

This makes it a trivial task - and means you can use cool web-based visualization tools (D3, etc), and combine with other useful APIs.


I think most scientific apps written in Python rely on native non-pure-python (written in C, Fortran etc...) modules which I'd expect to not work in this context.


If you wanted to go crazy you could use Emscripten but I could easily see pushing the C+Python parts to a server-side web service and then using Pypy.js for the rest.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: