AI-USE-STATEMENT-Long-Li.txt

CSE 132C Project 2 Spring 2023
Language Model Usage Statement 

First Name: Yunfan
Last Name: Long
Student ID: A16665173

First Name: Muchan 
Last Name: Li
Student ID: A15887258

Did you use any LLM-based tools such as GitHub's Copilot or ChatGPT during this project?
Yes

Tools that you used (enter N/A if none):
ChatGPT, VSCode C++ Plugins

Provide description of how you used it\them:
We used the plugins to help remind us some portions of the code on the fly, in particular
quickly navigating class methods and attributes. We used ChatGPT to help us understand
some of the concepts and syntax mistakes that we found confusing, such as generic member templates. 
Again, as we are not very familiar with C++, We also ask ChatGPT to generate some
code implementation by giving specific instructions with detailed logic based on our understanding 
of the project statement. This does saves us from the trouble of getting started, but trapped us with
other troubles: the code generated is very buggy and error prone. As a result, we have to debug 
from scratch--primarily chasing the pointers and memory leaks--which, while painful, greatly reinforced 
our understanding to the structure of BTree index. 

Did you find the tool helpful?
The plugins made code-writing super convenient and ChatGPT helped us converting our logic into code
and lifted us over some of the hardest learning humps.

Did using the tool detract from your learning experience in any way?
At first the tools made the programming feel a lot easier, as they accelerate the speed of 
writing code by generating raw versions of code. However, since we spent 
most of our time reading the documentations and debugging, in particular chasing the pointers, 
we were still able to formulate solid understandings about the project's objectives and 
details. In particular, from using the built-in GDB, we gained a deep level of
realization of the mechanisms of the B+-Tree Index and how the recursive call stacks are 
helping us getting the job done, while of course, still leaving behind a few notable mistakes
with respect to absurd error handling that we have to push off to future work because of 
our lack of programming maturity in C++.

Special Note:
Our code can't run consecutive tests (i.e. test1()->test2()->test3()->...). However, when we
comment all tests out but one, and run one input option at a time, for example, test1() only, and
make->./src/badgerdb_main 1, then make->./src/badgerdb_main 2, then make->./src/badgerdb_main 3,
all instances would pass. Similarly, then comment all out but test2(), do make->./src/badgerdb_main 1, 
then make->./src/badgerdb_main 2, then make->./src/badgerdb_main 3, all instances pass again. We would 
greatly appreciate it if you would please consider testing our code via this approach as we couldn't 
seem to resolve the issue that interupts the successive test runs. We understand if we are docked 
some points for this.