CS 371p Fall 2026: Dhruva Kumar

One concept from this week that challenged me was the lvalues and rvalues. It took some time for me to take the concept of an operator, such as “+”, “<<” and others, and actually parametrize it as something that can take in operators of a certain assign ability or strength and spit out other operators. Once I have thought about it more, though, this C++ ability is super cool! While I haven’t defined any of my own custom operators in C++ code quite yet, this is a very powerful feature of the language which I should experiment with.

I thought it was interesting how the designers of C++ used the lvalue and rvalue concept to add abilities such as return from += that are not available in other languages such as Java and C. I am not sure how much I’ll use that feature though. I prefer to have very simple lines of code that are easier for my brain to process than a master code line that accomplishes the job of 15 through masterful use of C++ features.

I am still working on applying the elephant and goldfish methodology. It has proven useful in this class and personal work using agents to build simple automations for me. With how capable the models are (such as the new GPT 6 Astra, see my tip below), getting the specification right is more useful than ever. If I get it right, the task can often be completed in one turn, saving the precious usage quota. If I am ambiguous, the model, well intentioned, will likely not do what I want.

The programming challenge this week was harder than I expected it to be. I thought about it from the wrong direction - processing numbers individually instead of the range as a whole. Once I went to help hours and was pointed in the right direction, things began to fall into place, so to speak. The problem approach is something I will have to add to my programming “toolkit” when I am faced with processing a stream of numbers. I am therefore glad to have been challenged with the problem!

My tip of the week is to try GPT 6 Astra. It is the latest frontier model from OpenAI, and it is excellent. The LLMs have been getting more and more consistent and accomplishing long running tasks. Even though I was extremely impressed with GPT 5.6 sol, Astra still is a nice improvement!

Dhruva