8
cb219
4y

Fuck, I knew that my code for my thesis would at some point become bad and very unmaintainable. Workaround here and there, everything put together "to fix later", just to make it all work "for now". I know what my code does where and when but my tech debt has reached a critical point, where a new idea and new procedure cannot be simply be added. Well, time to refactor and modularize as much as possible😪
Wish me luck that the whole project doesn't brake. Oh and of course so many different changes that I don't know what to put in git and in which order to do so.

Comments
  • 5
    @sqlkid my code will be part of the lane detection in a little autonomous robot car. My professor also told me that it would be ideal if my code could be used by future projects that rely on my results.
  • 3
    Does it have unit tests?

    I'm working on mine as well and I'm bouncing between "making it work" and "having nice code" at this point, although I still 100% test everything because I'm terrified of other engineers using my code and finding bugs
  • 1
    @alexbrooklyn no and I don't think its worth the effort (yet)
  • 2
    I hope you have unit tests!
    Makes the refactoring process much easier
  • 2
    @cb219 may god help you then :p
  • 3
    Create a new branch called "tech debt" and merge it in one massive commit. Not that anybody will look through your git history.
  • 2
    @sqlkid Hold on, it's a uni-driven thesis, no company included. 😅
    The project would be used by future students, no economics behind it.
    Let's say a future student does a thesis on steering control. He will need my results in order to make decisions or might have to modify my code in order to receive the info he needs.
    My professor is a great expert in these topics and knows that the project boundaries are limited.
  • 2
    @sqlkid It's about a competition between Unis to create the best possible autonomous f1tenth race car.
  • 1
    Commit everything - no code left behind!
  • 0
    You're writing lane detection software to possibly be used by future autonomous vehicles and you don't have any unit tests.
    That is the scariest thing I have read.
    Also highly negligent.
    I don't know who's more insane, your professor for wanting to promote this, or you for not doing it properly.
  • 1
    @dan-pud massive refactoring is on its way to hopefully satisfy SOLID ...
    I guess I might also introduce tests, but gotta keep an eye on deadlines 🤔
  • 2
    @dan-pud ... fast forward a few days, I can't believe I'm actually writing tests! Even more that they've helped me find some edge case bugs. Soo ... thank you! 😅
Add Comment