3 ms·
That's literally the opposite feedback of the rest of the world. There is a reason the scientific community standardized on Python, and bindings for all sorts o
by toyg 4y ago
That's literally the opposite feedback of the rest of the world. There is a reason the scientific community standardized on Python, and bindings for all sorts of things from Blender to KDE were more successful on Python than other languages.
- woodruffw 4y agoMy understanding of the "why" for Python in the scientific community is that it's twofold: it's very easy to get started in (and slowly build up expertise in), and (2) that a key set of packages (pandas, numpy) were created early and obtained critical mass. But neither of these implies that the CPython API is exceptionally good: (1) can be an overriding factor during the implementation of (2), and (2) means that the average scientific Python user doesn't actually need to touch the CPython APIs that much (all the work is already done!). So I don't think my feedback ("Ruby extensions are easier to write than CPython extensions") is actually incongruous with the rest of the world; the rest of the world picked Python because it's a better language in the ways that matter.
- toyg 4y agoMaybe, and maybe there was also a timing issue - in the sense that at the time it mattered, which was about 10-15 years ago, the Python api was better than most alternatives.
- woodruffw 4y agoSure, that seems likely to me as well! Edit: Well, maybe 20 years ago -- 10 years ago, Ruby's C API was about the same and still (IMO) better.