I want all schools teaching Java to complete beginners to stop and change to Python. The amount of boilerplate to start up a Java project hinders understanding of the essence of computing.
Below is hearsay.
Reserve Python for utility scripts and data exploration. For production, pick Julia if you are doing data analysis and AI, and Java if you are writing a service. The extra speed can keep your monolith fast enough longer so your scale wouldn't force you to split it up.
Yes Julia for data analysis an AI for speed (apparently, still not found the time to play with it properly yet), but with the tradeoff being that Julia is still far from an 'accepted' language.
A significant proportion of the data-science field that I interact with I would doubt have heard of it. FWIW when asked about why Jupyter notebooks are called Jupyter, I would give it 50:50 that a user would be able to state it stands for Julia-Python-R notebooks. Looking at AWS as a major example, their docs for their data science services talk about python first, R sometimes, and Julia extremely rarely.
That said, I am not embedded in the Julia community at all, so there is significant bias in my understanding. Is Julia something you have put into prod? What were your experiences, would you mind descirbing a high level over view of the tasks and the stack? I'm always keen to learn.
Sorry I have to disappoint you, but as I wrote, it was hearsay -- I am not involved in prod system with Julia. TIL that Jupyter is Julia-Python-R! I've read a few cases of unintentionally running loops in Python instead of numpy, which is slow. I hope you have better luck exploring Julia further elsewhere!
We're a place where coders share, stay up-to-date and grow their careers.
We strive for transparency and don't collect excess data.