Interesting, useful and funny read by @jgbbrd on Stacked Diffs Versus Pull Requests jg.gg/2018/09/29/stacked-dif… Lots to learn about good practice even when using PRs. Rebase interactive FTW

I've had to try several times to read the article due to the number of outright lies and miscommunications in the first few sections. In the end I had to stop because: whether or not "Stacked Diffs" are any good, the supercilious ultracrepidation overruled any possible value.

Could you be more specific for my own benefit and education? :-)

First sentence of the fourth paragraph of the next section about how engineers don't work on easily broken down work just suggests that author's workplace doesn't do a good job of work planning and this leads them to make a sweeping statement which is false.

I'll admit that beyond that, I just give up. There's an assumption that engineers are unable to operate Git efficiently, which is sad, and that by piling up loads of semi-unrelated changes on one branch you get something you can confidently submit for review which seems odd.

All in all it's very sad because the article has some good points related to producing integral changes which can be reviewed as a whole. But it's couched in so much sensationalist nonsense that the message gets lost.

What I've learned after Codethink is that we were very good at software purity and some processes. You can't expect everywhere else to be the same. Codethink is our own bubble of opinion, but not that many people care about using git properly