3 ms·
Too bad it's destined to be abandonware. Just think of all that money and engineering time that could have been spent working on CPython upstream, rather than a
by throwawaaarrgh 6y ago
Too bad it's destined to be abandonware. Just think of all that money and engineering time that could have been spent working on CPython upstream, rather than a fork that'll die when the company gets bored with it.
Aside from faster CPython, I'd love to see a PIP that standardizes module naming convention & inheriting+extending existing modules, rather than completely reinventing the wheel and making a totally new module just to add a single feature that some other module lacked. That plus the new regex extensions and Python will be almost as useful as Perl. :-)
- minedwiz 6y agoThe readme does say that the goal here is to either get this work upstreamed to CPython or to inspire/inform improvements to it, not to replace/supplement it.
- throwawaaarrgh 6y agoBut it's probably not going to go upstream. And probably very little of it will go into other projects. Every time one of these forks gets dropped on the internet, nobody does anything with it, because it goes out of date shortly after it's dropped. Probably a couple specific optimizations will get picked up from it and put upstream. Versus if development had started and stayed in upstream, you wouldn't have to force people to pick through a huge patchset looking for things they personally want upstreamed.
- nuclear_eclipse 6y agoWe have multiple CPython core developers at Facebook, and the Cinder team (among others) are heavily involved in discussions around CPython's direction. There are many pieces of Cinder that would never get accepted upstream, or would take a decade of PEPs and incremental improvements to get there. Cinder is both a proof of concept that these ideas can work in CPython, and something that we can use today, rather than waiting a decade to realize these benefits from a pure open source perspective.