Learnings from Development

It’s not yet finished, there’s lot of work to do.

However, there’s always learning to do.

As stated, originally this started as a personal python script and grew rapidly.

I gave myself two weeks for a personal sprint to do as much as I could outside of looking for a job.

Some of the bits are cobbled together from examples/docs, it’s why the returns from some functions are inconsistent. I want to be clear that there’s no AI in any of this, whilst it’s a useful tool for some things, I think having it paste over a knowledge gap just leads to problems when you can’t explain why something does what it does and you’ve hit it’s limits.

Having got this far, what should I do differently?

• Start with the API and docs
  • Technically I did this, it’s all based on GET/PUT etc, with data models and follows REST
  • However, adding Marshmallow and the API docs has proven a little incompatible (Schemas vs Models etc and duplicating data), this is mainly a skill issue on my part and an assumption that I could bolt it on top.
• I started with Blueprints etc, and it’s all modular, apart from Core, which should have a clearer designed boundary on it’s purpose as there’s a couple of leaks
• I know Python and SQL, Docker, I’ve made build pipelines, but I didn’t know that much Javascript, much about Alembic or any Bootstrap stuff, that’s been a lot of tech all at once
• I’m a terrible web designer, those people work magic

Thing’s it’s re-inforced in me:

• I genuinely love having a problem to fix with tech
• End users can break anything, leading them without being there is still an incredibly useful skill.
• MVP is still vital, there’s stuff in here that doesn’t need to be, this was originally a 2 week best effort sprint project,