Software, and the Overstated Effectiveness of Authoritarism.
I've been thinking a lot lately about the parallels between these two
There's kind of a prevailing view in software that comittees are where software goes to die.
- Singapore is help up as the model of "Doing it right"
- We can draw parallels here to the Python BDFL
The problem with this approach is that mis-states what some of the effective points of democratic institutions are
- They're durable, projects that break away from the main development cycle with special autonomy have an unlimited mandate for a short period of time. The problem is that when those charismatic leaders die (change jobs) they leave behind all the problems they create, because there is no system there to look after them
- Consensus is built early, there's the idea that everyone is trying to slow you down, and that if you were just given the free roam to do what you want then it would be easier. But we know in software development that software development
- Survivorship bias exists in those authoritarian projects, the python example is famous and successful but the things that come second and third to mind are those that didn't work, but they're not considered as failing for the same reason. This means that there are rose tinted glasses that "It will work if we do it"
- Democracies have an image problem, there are enormous sofware successful projects. But they get less credit than oher types of projects, because they come with the reputation of slowness and ineffecicies
Democracies are inefficient, but the best