Who said I "changed" to Python? I use C++ (rarely C# these days) , I use Python, Rust, R, and a number other languages, each as I see fit. Your post that you believe argues against python could have been written by an angry and clueless 7 year old.
You use C++ and you didn't know about JetBrains IDEs? It has the best C++ IDE available, Clion. You would know that if you were a true C++ developer.
Of course, only you know what is best and nobody else. When being asked what makes your suggestions better than other choices your standard answer to me and everyone else has been a consistent "because my choice of ide and language is best, and now fuck off". I am absolutely convinced by now that you are a 100% poser and pretender and actually know very little. No real software developer or architect will corner himself into a single language and IDE like you do. Feel free to respond or not but in my book you disqualified yourself from speaking on this issue altogether.
As I said, you are funny. Let me ask you one simple question: what is the language I use the most for trading?
I am not a programmer as my discipline is in industrial engineering. I learned Fortran, PL/1, Cobol, Pascal, C and Simscript during my undergrad. A university friend introduced to me perl. According to him, the financial institution he was retiring from had already started migrating from perl to Python. Is perl dead? Nope. By the way, Strawberry perl is freed. All my data crunching is with perl, and some protyping in Excel VBA. If a file has more than 1.048M lines, then power pivot is required. If it's a few million lines, then perl script can handle with east. Of course, MYSQL can do it but database not that structure. Since I am using Amibroker for charting, screening, and backtesting, AFL scripting is a must. I also use Windows Script Host i.e. Jscript for OLE Amibroker automation. All programming languages are designed for specific purposes. There is no good or bad programming languages. Even assembly has its purpose.
That's what an engineer would say and is absolutely correct. There are preferential approaches and solutions, but often times those preferential approaches don't get taken for various reasons : budget constraints, domain specific knowledge of existing resources such as engineers and developers, legacy code that just works and has worked for decades and better not be touched, and then decisions by heads or supervisors. Very rarely do developers have a choice in non-retail space. But in retail space the choice is up to each individual and again here there are metrics that inform what language choice tue individual makes. Excellent python backends that run for 10 years straight without ever breaking can be created and elaborate type safe oop language based solutions can be had as well. No developer worth his/her salt will ever rely on a single language out of personal feelings and emotions and stomps on everyone else's preferences.
Sometimes, simple question is very difficult to answer. For example, what will be tomorrow's stock price ? Just a simple answer, up or down.