What I read into this is a feeling I have to fight in my own work: it's easier, in an emotional sense, to fiddle around with your tools than it is to face the r

What I read into this is a feeling I have to fight in my own work: it's easier, in an emotional sense, to fiddle around with your tools than it is to face the real problem. The amazing things you're imagining creating don't really need a replacement for std::vector to succeed -- people make great art with all sorts of tools way worse than the STL. The way I fight this urge in myself (and balance it against the real need to occasionally improve a tool or library) is to occasionally make a project where I make a point of doing everything the most straightforward/simple/least-clever way possible, even choosing the hackier way when available as it makes me uncomfortable. I find it helps me reset my acceptability meter. Usually I find the project works out just fine, and all the things I imagined needed improvement were just avoidance.