Agree with this. Shipping is the most important quality. The faster you ship the faster you would be enjoying anything
There is something to be said for the good old days when we modified PHP files in an editor over FTP :)
Never done it professionally. But developing in production would make me ship faster, and using php would probably limit my tendency to make gold plated solutions.
Code and software quality was surprisingly better than it is today.
Today? In a process heavy company with Jira, PMs and QA it can become a week-long affair, unless I (the lead developer) bypass parts of the process.
The perfectionism is in the process, but the outcome is not quality. Come to think of it, perfectionism never really leads to anything of quality.
TeX stood the test of time and it was released as close to perfection as it gets in non-life-critical software. (One could argue it wasn’t perfectionism at work, but sound top-down design.)
It even leads to more different changes and more ground to improve upon again.
So.. Can you trick yourself into directing the perfectionism into shipping fastest?
At one point, the thing I wanted to learn was "call something done". So I did that, and because it was a side project with learning as an explicit goal, I was able to actually do that - it was literally impossible to ship something imperfect, as shipping would already mean I'd perfectly reached my goal.
But that turned out to be enough faking it to make it - the feeling of shipping it felt good, and gave me the confidence to spend a small amount of time to polish my side projects up enough for shipping, rather than a large amount of time to do everything I want. And it's way more fun now to be able to occasionally pull up an old side project of mine that's still usable, because I got it to a usable state in the first place.
(In case anyone's interested, you can see the project, and how small its scope was, at https://agripongit.vincenttunru.com.)
Perfectionism comes from iterations