[mark_l_watson]

I downloaded a free trial of Mathematica about two weeks ago. I appreciate all of the built in data sources and libraries, the notebooks work well, but it is hard for me to go all-in on something that is not an open source language.

I am a Lisp/Java/Python/Ruby/Haskell programmer, but, the language I really wish I knew inside and out is Julia. I think that Julia is hackable like Lisp but I just don’t have the experience to do it. I would not be surprised if Julia in five years becomes an open system like Mathematica and the Wolfram Language - with lots of libraries supplying useful data, lots of libraries and language extensions, etc. sort of what Python is today but much more efficient and better for building very large systems.

[giancarlostoro]

Julia is amazing its the language I wish I had time for too. You can literally have it print out the assembly for the code you write. I know C and such can do so. But Julia does it within its little REPL tool. It is incredible.

Unfortunately I love languages like Python and D the most: multiparadigm but not crippled by one single programming approach like OO or Functional. Just do what is best for you and as you learn the language it gets way better.

I would pay good money (if I had good money) to invest a lot more into D. It has so much more potential imho. I would love to see a UI library in raw D that comes out of the box with D in the std lib for example. Same with an out of the box web server like Go does. Lastly maybe some editor like IDLE for D. Coded fully in D with its own UI library like Racket does. I also love Racket it makes me feel like I can do anything and indeed I can.

GUI programming is not emphasized enough in new languages to the point where Electron took off cause when the UI is just HTML and CSS mostly, the sky is the limit. You can reuse skills you know! Qt even has some Electron like stuff going on which is neat.

[ssivark]

Julia is a fantastic language and ecosystem, but I don't think it's the right thing to compare Mathematica against. As a language, Mathematica's computational model and strengths are closer to Lisp, IMHO. As a "computational system" their goals might be similar, but I can't fathom what Stephen Wolfram actually aims at, for Mathematica.

Julia, btw, is actually a language you could potentially know inside out. The discipline to keep the language simple enough that LLVM can generate fast code means that it's not hard to understand and there aren't many gotchas. On top of that, the community discussions (on Github issues and the Discourse forum) are very interesting and enlightening to read through.