while knowing your tools is always a good idea, i disagree with the overall notion of "don't divide labour" because a lot of the real-world problems software ha

while knowing your tools is always a good idea, i disagree with the overall notion of "don't divide labour" because a lot of the real-world problems software has to deal with are fractally complex. for instance, to pick a seemingly well-contained problem: if i wanted to write a collaborative text editor i would absolutely want to rely on someone else's work for font rendering or fault-tolerant networking, because each of those are problems that i could spend years on, and still not have covered all the failure modes and corner cases.