@StefanKarpinski, very enlightening writing.
I really like the fact that Scientific Computing is a first priority.
Going forward, do you think it is achievable, in the future, to write code in its Vectorized Form + Decorations in Julia and have performance of highly optimized (SIMD + Threading) C Code?
I liked the guide line that what sacrifices performance won’t get in.
Python, in my opinion, went too fast with its flexibility and types of variables.
Sometimes its code looks like Black Magic.
I prefer language with less built in types which sometimes means you code more but also the code is simple to grasp.
I hope Julia will remain simple in its Base.
@John_Gibson,
I think Julia’s charm is how deep it lets you go.
For instance, @inbounds macro, Choice whether to have SIMD or not, use Multi Threading or not, etc…
It means that on the surface you have the usefulness of MATLAB like language yet when you want to speed things up and get your hands dirty you can do it within the same language, Julia (As opposed going C in MATLAB).
Yet still, the challenge of Julia will be to close the gap of performance when writing Vectorized “MATLAB Like” code to what can a user get when going down.
Hopefully it is doable and if it is doable, Julia seems to be the best candidate to achieve it.
Thank You.