The interviewer briefly asked about what I do currently. And he asked the reason for switching. I was asked, what does this statement print: printf("%d", 4.8); I asnwered "4". I was then asked, who does the type casting "printf" func or the compiler or the clib library. I was not sure, I asnwered printf. But after the interviewer did some explaining, I admitted that it must be the compiler. What does static mean? I answered that it allocates variable in static memory, and it limits the scope of variable to that cpp file. Does c++ add any other meaning to the 'static' keyword? I answered about the member functions being static and that they can be called even without instanciating any object. For example, String can have a static split function. What is 'Mutable' keyword? I explaind that mutable is used to allow a member variable to change even though the func is a const. What is a usecase of 'mutable' keyword? I tried to explain but ended up with unclear mind. I couldn't think of any usecase at that moment. I told I am sorry, can't think of a usecase. Explain type casting in c++? I enumerated: dynamic_cast: explained dynamic_cast that it returns null if we try to cast an incorrect type. const_cast: Const cast is used in place of mutable keyword. It should be used as a last resort. reinterpret_cast: I told that it is used to forcefully type cast object of a different type to another unrelated type. static_cast: (I told that I forgot) Interviewer wanted to get into detail of dynamic_cast. The voice was breaking, so it took several minutes to understand that he was just asking if a object is of derived type pointed by the base type pointer, will we be able to dynamic_cast. I said 'yes', we will be able to do dynamic_cast. I told that dynamic_cast needs RTTI to know the type information. I think, he basically wanted to hear RTTI; I told that RTTI is enabled by default in some compilers and sometimes you have to explicitly set the compiler flag to enable RTTI. Data structures: I was asked which data structures I have used. I mentioned linked lists, maps. How is map implemented? I said using hash tables. How do we implement hash tables? I couldn't answer. We have 2 linked lists, we have to find out if both of them merge together in the middle or not? I asked a few questions, can we mark nodes? NO. Do we have to tell which nodes they merge at or we just have to tell yes or no? I was told that we don't have to tell which node they merged at, we just have to tell if they merged or not. I answered that we should traverse both the lists till th end and see if the last node has the same address. It was the correct answer. Network programming: Difference between TCP and UDP? I explained the differences: 1. TCP is connection based, UDP is not. 2. UDP is faster in higher asynchronus traffic. (especially in traffics like that of stock markets). 3. UDP is less reliable, TCP is more reliable. I was asked about tftp protocol, what does it use TCP or UDP? I answered incorrectly, I said TCP. The interviewer corrected and and said that TFTP uses UDP. But he asked why? I went in circles but couldn't answer in the end. The interviewer asked if I have worked on multithreading. I said yes. What is the difference between a Process and a Thread? I explained how threads reside in the memory space of a process. Why use threads at all? I said because threads consume less memory. Then he said, why not use threads all the time, and why use any process at all? I said because then it would be a security issue. OS gives every process its own virtualized memory so that no process can access other process's memory even by mistake. How do threads communicate with each other? Synchronization objects: Mutex, semaphores and Critical Sections. Can Processes communicate with each other? Yes. Using memory-mapped files, pipes, network. Then before ending he asked if I know XML. I said yes. And then he asked me about DOM and SAX. I explained that DOM consumes the entire xml in memory before we access it. While SAX read XML and generate events during reading it. That DOM is useful for smaller files. and SAX is useful for hugh files (which we don't want to store in memory). I got an email in couple of days stating that I might not be the perfect fit for them. Bloomberg has a very high rejection rate, so i guess its okay for me to be rejected, but I can consider that experience as a practice for interviews with other companies.