Cinder is fully compatible with the CPython's C-API

> We don't have the capacity to support Cinder as an independent open-source project

Our team is smaller than you might think! You also left off the last half of that sentence, which is quite important: "nor any desire for it to become an alternative to CPython."

We'd love to see the ideas and implementations in Cinder end up in upstream CPython in some form or another. It'll make our lives easier in that we won't have to maintain as much forked code, and hopefully everyone else will benefit from a faster CPython.

> Our team is smaller than you might think!

I thought so but the other half of my brain was screaming for "FB has so much power, how come they don't invest more in open source".

But well, that's economics at play I guess. I'm an idealist after all... Thx for replying !

A couple of the folks who wrote MonkeyType (me and _carljm) also work on Cinder :)

Good call out! I’ll make sure it ends up in the readme.

> Maybe there could be a subset of Python that removes the hyper dynamic parts so it can be compiled to something faster.

What you’re describing is essentially Static Python (mentioned in the readme)

We’ve chatted with kmod a few times and let him know we were open sourcing Cinder. Hopefully the projects can learn from each other. As _carljm mentioned, Pyston was restarted way after Cinder was in production and we had already implemented a significant amount of the core JIT functionality.

Envy was painless for me (if you need to get an nVidia card working): http://albertomilone.com/nvidia_scripts1.html

I participated in this study at Stanford for a while. The results are pretty dramatic. I doubled the number of pull ups that I could do in six weeks and went from being able to do 12 reps of 155 on bench to 12 reps of 175.

Looking back at my own logs from three summers ago, I see it took me from June 5 to July 17 to go from 12 pullups to 24. It seems like your rate of improvement is about par for pullups and a little below average for bench press. I doubt the glove made any difference.

That's very good progress, but I'm not convinced the glove is what's responsible. That sounds like reasonable gains for someone who is working out regularly and trying to improve. Good, of course, but still within reason.

Valgrind is also incredibly useful if you are working with Linux.

Yep, especially with tools like kcachegrind for visualizing the output of valgrind's profiler (valgrind --tool=callgrind)

Um... what? You install valgrind by selecting it from your package manager (e.g. "apt-get install valgrind"). You run it by prepending "valgrind" to the command. Then you read the output. Valgrind is one of the most dummy-proof development tools I'm aware of. For someone to claim that they don't have the patience to "learn" it is just beyond me.