News for hackers. Stuff that matters.
Re:Show HN: 1M rows/s from Postgres to Python by 1st1 on Friday August 05, 2016 @12:08AM (#12229985)
I use SublimeText with a Cython highlighter; @redcrowbar uses Atom. I started coding uvloop & asyncpg directly in Cython, basically from scratch.

Debugging & refactoring are pretty much as easy as for pure Python code (which is a huge win over C). Sometimes when you're writing low-level code you can get a segfault, in which case lldb (or gdb) help you understand what's going on relatively quickly.

Re:Show HN: 1M rows/s from Postgres to Python by denfromufa on Friday August 05, 2016 @01:08AM (#12230168)
So do you use cygdb to debug Cython code or directly debug the generated C-API code?
Re:Show HN: 1M rows/s from Postgres to Python by 1st1 on Friday August 05, 2016 @01:16AM (#12230200)
I usually debug the generated C code directly (usually it's quite easy to understand the bug once you see where the program crashes.)