I can't say that I like or use Julia, but just regarding your comment, and because I've seen similar comments about other languages around:
> 2. Julia uses an "end" keyword everywhere, which is imho too verbose (and the corresponding "begin" is missing so it's inconsistent).
This is an absolutely childish approach to comparing or selecting programming languages.
In my eyes, it says a lot about the maturity of software development as a discipline that a big chunk of our debates are at this level.
We should discuss about quality, breadth and depth of standard libraries, quality of implementation of the most common interpreters/compilers, etc.
I don't want to fault you personally, OP, I think this approach is quite widespread, unfortunately, one could say that it's part of our software development culture at this point.
While you do interface with it all the time, it's not something you'll be thinking all the time, at some point it will be completely invisible even though it's there (unless you're that bored). What will really be a big part of the experience is what offers actual resistance to solving your problem and actively waste brain power and not muscle memory, be it the fact that language semantics forces you to write the same thing all the time when a particular feature (like macros) could handle it trivially, the lack of some particular type safety makes it so you keep losing time debugging the same error, the interactive tools being lacking forcing you to waste time debugging with prints, you have a workflow where the JIT lag keeps breaking your pace, the community doesn't have a culture of documentation making it so you'll lose time trying to decode the source, etc...
> 2. Julia uses an "end" keyword everywhere, which is imho too verbose (and the corresponding "begin" is missing so it's inconsistent).
This is an absolutely childish approach to comparing or selecting programming languages.
In my eyes, it says a lot about the maturity of software development as a discipline that a big chunk of our debates are at this level.
We should discuss about quality, breadth and depth of standard libraries, quality of implementation of the most common interpreters/compilers, etc.
I don't want to fault you personally, OP, I think this approach is quite widespread, unfortunately, one could say that it's part of our software development culture at this point.