If anyone has some kind of study or more concrete example, I'd love to hear about it.
I also think it's highly likely the anecdote is supposed to be illustrative rather than literal. IMO it's meant to be advice to students not to fret about how perfect their work is and that with time it will improve. It's just been distorted by the "one quick hack" culture of self improvement.
If it is a made up anecdote, that would be rather misleading under a headline that says "science-backed".
The article argues that it's supported by data that focus on quantity leads to better quality than focus on quantity.
It doesn't argue that it's merely a compelling idea that we should try for ourselves.
Plan 9 and GNU Herd vs Linux.
Spring vs EJB.
Rest vs Soap vs CORBA.
C vs Algol.
It's easier to do something simple and iterate on it than to do something perfect from the start.
Rather the whole point here is that when learning it's better to focus on quantity over quality AND by doing so that quality will naturally be better. The parent is looking for examples of this latter idea.
But let's take the idea to the extreme, imagine we are building an system, one team starts building and improving on the design for 6 months. Other team builds and starts anew every 2 weeks. Who would have a better system at the end of 6 months? Tough to tell, the iterative one will probably build it from the ground in a better way with less technical debt but who will have a more complete system or more features.
This is basically agile vs waterfall.
EDIT: Ah, you mean the order is different :)
- C++
- CMake
- PHP
- Postscript
And it can work well:
- TeX
- Git (not the later additions)
But this thought experiment doesn't tell you what works, as either outcome is quite compelling and plausible. So it's not much use as a thought experiment, other than to show that it might be interesting to study it in the real world.
I.e. it might be "obvious" to you that focus on quantity over quality leads to better quality as assessed by another in the thought experiment. But it might be equally "obvious" to someone else the other way.
A real study would be interesting, because there might be a real world consistent pattern. It might actually be useful to know which one works best for real.
Also since this human behaviour we talking about here. Experimental Psychology is a field known for its credibility crisis, its inconsistent and possible non-reproduceable results, that I have not much faith in it to give us useful insights, and I would be more interested in hearing many of those matters addressed from a purely theoretic point of view, than through conclusions from dubious experimental data.
