Hacker News Reader

Comment by Zuph | Hacker News Reader

Processing power is only one part of the equation, though. These Cortex processors are significantly more constrained in RAM and storage than computers of 1995. Run Python on a 100MHz Pentium with 192 KiB of RAM, I'll be impressed.

View on HN

Python was ported to PalmOS, where it could run on a 16MHz CPU with 64KB of usable heap space: http://isr.uci.edu/projects/sensos/python/

That's golden news! ("news" in a sense "new is well forgotten old"). I heard about bunch of previous small Python implementations, but never about that one. As a MicroPython contributor, I'd be proud to think of it as MicroPython's spiritual ancestor. It is even an 1.5.2 port, wouldn't think someone attempted to space-optimize CPython (MicroPython is developed from scratch, CPython3 hardly could be shrunk into that).

For you reference, MicroPython can start up in as little as 2K of heap. Of course, one hardly can execute something more than a simple expression like "1+1" in that memory, but 2K also include memory required for interactive interpreter (and parsing into AST and compiling into bytecode).

Palm OS3 indeed (1998). I didn't see Python on Nokia, PyS60 till 2005:

- S60 https://en.wikipedia.org/wiki/Python_for_S60

- S60 http://www.onlamp.com/pub/a/python/2005/04/14/s60.html

- N73 https://rubayeet.wordpress.com/2009/01/30/installing-python-...

I never did see any real usage aside from one of the flickr devs using an N75 to upload/query flickr via Python.

Reply on news.ycombinator.com