Novice CL user writes mysql access in CL as an exercise

I've been guilty of exactly that (in java, a persistence layer for classes) and was rightly shown why that is not a good thing to do. Too many edge cases and a serious maintenance problem in the longer term.

Before you know it you are maintaining your database connector more than you are your original reason for writing it (the problem that needed to be solved).

I'm not too familiar with the common lisp ecosystem but there is a reason why there are N reasons of doing things there, all of them with varying degrees of feature overlap and stability.

Unless you have a really good idea on how you are going to move the field as a whole forward by re-implementing something your time is likely (but not always!) better spent by improving a solution already out there than by writing another partial implementation from scratch.

The first few weeks of progress on anything new it always seems like you're literally on a roll, after that the slowdown comes and the hard work begins.

Perhaps there is a point it's so easy t build useful stuff that the ease builds fragmentation. If ít's easier to write a new library than to adapt to one already existing, you have a situation where everybody just writes new libraries from scratch. The easier to write, the more fragmentation you get.

Keeping fragmentation low requires effort.