[mewpmewp2]

To me in a lot of cases it seems even as a reviewer, it's harder to understand big picture if someone splits up the PRs.

But as a code writer myself, for example, if I am building a new feature, firstly it's really hard for me to know what the whole thing would look like without going through it all and it's probably very iterative process as I'm doing it, so I usually wouldn't be able to split it up or it would very suboptimal to split it up before I've finished everything.

Then I would try to split it up as I've finished to appease reviewers, but again, it requires whole lot of creativity to do. Should I try to split up shared component first? Because I surely can't split up whatever is using those shared components. If I do then, people won't see whatever is implementing those shared components so they won't have understanding on why those shared components provide certain functionality etc.

Overall it complicates a lot it seems because if I was to do it during my iterative process then I would write a PR, later refactor bunch of it anyway, and I would do it in the order that feels best for me, but wouldn't necessarily be easy to understand for anyone not within it.