3 ms·
It's a cool idea, but when you sick it on something that uses numpy, a very common use case that people would want, it basically dies.
by danmg 8y ago
It's a cool idea, but when you sick it on something that uses numpy, a very common use case that people would want, it basically dies.
- weberc2 8y agoSurprising since it advertises itself as "It is the extremely compatible Python compiler."
- metalliqaz 8y agoThere's a bunch of discussion and links in this[1] SO question. It looks like numpy's size becomes an issue. It appears that some people do use the two projects together, though. [1] https://stackoverflow.com/questions/37713589/nuitka-error-cannot-find-in-package-as-absolute-import https://stackoverflow.com/questions/37713589/nuitka-error-ca...
- danmg 8y agoEven if it the end it spits out an exe, it's 100s of MB. This makes it a non-starter for short lived numerical programs.
- Eli_P 8y agoAre you using --standalone flag? Nuitka scans for dependencies recursively, that's why output bundle's so big. I don't understand why you've that large .exe, a bunch of DLLs is what should be heavy.