8 ms·
I think he is making the distinction between 3 different categories that readers are in general lumping into the 'interpreter': 1) Python is executed by an int
by apalmer 6y ago
I think he is making the distinction between 3 different categories that readers are in general lumping into the 'interpreter':
1) Python is executed by an interpreter (necessary overhead)
2) Python as a language is so dynamic/flexible/erfonomic that it has to do things that have overhead (necessary complexity unless you change the language)
3) the specific implementation of the interpreter achieves 1 and 2 in ways that can be significantly slower than necessary
Seems he is pointing out that a lot of performance issues that are generally thought to be due to 1 and 2 are really 3
- gsnedders 6y ago> 2) Python as a language is so dynamic/flexible/erfonomic that it has to do things that have overhead (necessary complexity unless you change the language) As PyPy demonstrates, much of this doesn't need to have anywhere near the overhead that it does in CPython. You can absolutely do better than CPython without changing the language, and you can do better than CPython without a JIT if you start specialising code.