Coding in basic. Are these people 90 years old? Almost eveything modern is C based, you might not actually code in C but unless you have some ancient legacy systems I dont see how someone today can get away from that dynamic.
Python is slower at run time, but it is much faster for iterating over ideas and building new strategies. Components of the system that almost never change (data handlers, fix engine) are best build on C or even on a FPGA but python adds value in areas that are constantly changing and require fast coding time.
If you want to run 1 yr worth of ES futures tick data through your strategy, you are talking about a few billion ticks/bids/asks. When you try to pass each of those billions of data points through your strategy, the choice of your language matters. I have a C++ implementation (btw: my specialization is c++ low latency) that churns all the data against my strategy in about 20-25 min... a python implementation of the same would take forever and very hard to debug... I however use Python for all my other usecases that don't require speed... for example, I have python scripts that crawl certain websites and distill certain economic news information. I don't use C++ here; thats VERY important data which helps my strategies decide how to place trades... and ... soooo... use the correct language for the correct usecase... as a professional programmer, you should never get biased by your preference (I myself have problems following this ) but, this is important! my 2 cents; hope this helps.
Nice videos and good to know some of these things. However, you rarely ever get to use such theoretical knowledge, until you build libraries, distributed/fault-tolerant DBMS/OS'es for others to use, or simply to understand why something doesn't work as well in production as it did in testing. It's funny hearing theoreticians talking about programming, because their natures are inherently anti-programming. Such people seek the grand, generic and unified descriptions of the world (constraints), which is not what programming is about. Programming, if you think about it, is really about the execution itself and all of its explicit details, which is why in the end OO is so tempting, but still a dead end (sorry C++11). It's very often mostly about the process (actions), being a good start and focus throughout the lifecycle. Also programs need to be as near as 100% correct as possible, sometimes just because we can, but more often because it's a pain in the butt to fix those 1-5% where it did go bonkers. In the real world, the last percentages never get fixed, simply because it's not cost-effective enough to design for it in the first place. In my youth, I went into the same trap. Built alot of libraries, thinking I was productive on my own hobby projects, while I was really just learning programming and creating waste (non-used code). Theoretical models are kind of the same trap, at least if you think you're doing programming, because spending time describing the world, is not really directly productive, but it may or may not advance computer science. I suspect, it just as well may be a way to psychologically cope with an inherently uncertain and undeterministic world. Lastly, and this is a bit funny, theoreticians are often searching in vain for "holy grails": Ie. to find the solution where the definition will tell everything about the function, without side-effects or revealing the inner workings and dependencies of the function. As if an object can be modeled perfectly, or is not good enough by itself (hello Lisp!). It's funny, cause it's kind of the same type of hunt when you see a cat or dog chasing its tail, and is the kind of activity where any experienced practitioner will just chuckle and mumble something about weighing the right level of granularity against acceptable costs and risks. The search itself is always respectable, but often a bit delusional all the same. Sometimes though, it sparks new light and understandings that changes the world forever.
Focus on learning c++ mostly. So far have done even guis in c++ but thinking about Python or something easyer also, it may be good for testing/refining before porting to c or for code that does not require high peformance.