Peekaboo Pairs

Give it a shot at Peekaboo Pairs

When I came across AWS Game Builder Challenge, I had multiple goals in mind with this, with the obvious being the prize money, the more realistic one being - see how far I can get to with the latest stack of front end technologies in developing a complete solution. I have been a backend developer for my entire engineering career working on C, C++, Python, Perl, Ruby. My indulgence in front end development is minimal and probably a decade old tinkering something with jQuery and primitive HTML, CSS. The other tangetial goal I had is to develop something that my kid has fun playing with.

My initial attempt for is a multi-player board game, spent few days on it, even though I am using the Code generation tools like Amazon Q, ChatGPT, Google AI Studio the sorts of problems I am encountering weren't easy to solve, I was not able to get to a workable solution with help of online documentation and LLMs, I realized it would require lot more domain expertise to proceed further with that, which I don't have at the moment, so I gave up on this challenge.

With 4 days left in the competition, I didn't want to feel defeated in this, so I came back to Amazon Q this time right from the beginning in terms of what game that I could try for this and picking a simpler game which has minimal components and I went more methodical with this one, starting with minimal skeleton and slowly adding features with increasing complexity, so that I can debug and potential fix thigns on the way. Fortunately, this approach worked, I am amazed at LLMs(in this specific scenario with Amazon Q) capability to writing code that helps in full end to end development.

There are 960 lines of code including all the files in the src and few configuration files at the root level. All these lines except 7 lines(in src/styles/GameBoard.css) including comments 😀 were generated by Amazon Q, all I had to do was click Insert At Cursor of the generated code in the right file.

A full chat history of prompts and responses captured at docs/amazon-q-full-chat.md for the curious. I wish I had an easier way to capture the marked down response generated in the chat window. Also included some images captured of the prompts and responses in docs.

I am quite surprised on how far I was able to go with the Amazon Q IDE integrated Chat plugin, highlighting some of the aspects of the Amazon Q usage below:

Couple of screenshots of the game: