26

Devs : Lets pick library X, it is well know piece of open source technology, actively maintained by community for over 10 years.
Architect : NAH, it is an overkill to use it in our project , lets build our own solution.
*2 Months later*
The code base is hundreds of thousands lines of code, we basically started to look at library X on GitHub to copy features or get inspiration from that code. In that time we delivered 0 business value, it is horrible to use it and we constantly adding something or bugfixing because no one thought about something in first place.

Comments
  • 1
    I feel you... What happened to the good ol' "Don't reinvent the wheel"?
  • 3
    @BixelPitch: Dependency hell happened. Selecting where and what dependency to use is a hard problem. It might be easy when it comes to heavy, well-contained stuff like PDF rendering.

    But if you initially only need to do some reasonably simple task X, you basically have to predict the future evolution of the scope of task X.

    I avoid pulling in additional dependencies when i reasonably can. So i do reinvent a lot of small wheels but avoid reinventing the really big and heavy ones. But i also never had to deal with nodejs-level dependency debt...
Add Comment