I don't read the screed as being against synchronous communication; just against synchronous communication being the default.
> Else you end up waiting an inordinate amount of time getting simple answers to simple questions
Well, this is the problem. Expecting an immediate response kills deep focus. You should have to wait for a response, for anything that isn't an emergency. Then the question becomes, what's an appropriate amount of time to wait for a response? And how can you a) measure the number of async conversations where ping pong resulted in conversations taking much more time than expected, and b) come up with blameless / systematic responses (e.g. internal wiki maintenance) to reduce that measure over time?
> Well, this is the problem. Expecting an immediate response kills deep focus. You should have to wait for a response, for anything that isn't an emergency.
This is a ridiculous assertion.
"Hey, I think we should change feature Y to do X instead, but this will have an affect on your feature. Can you absorb this into your workstream? Btw I'm writing this code right now and want to push a PR today"
... wait 5 hours because asynchronous ...
"Mmm, maybe. Could we make this slight tweak that will still get you what you want but be easier on me?"
... wait 24 hours because now it's sleeping time ..
etc.
Real-time conversations are extremely valuable. Let's not pretend like they're not.
Yes, waiting five hours to then have a synchronous conversation is one way to handle this. Another way is to not assign feature work asynchronously, but to have a synchronous meeting where the feature is assigned, to try and ensure that people are aligned before they go off and work on their own.
> Real-time conversations are extremely valuable.
Of course they are. It's the ability to have a real-time conversation on demand, with no consideration for the counter-party's focus, that is ruinous.
> Else you end up waiting an inordinate amount of time getting simple answers to simple questions
Well, this is the problem. Expecting an immediate response kills deep focus. You should have to wait for a response, for anything that isn't an emergency. Then the question becomes, what's an appropriate amount of time to wait for a response? And how can you a) measure the number of async conversations where ping pong resulted in conversations taking much more time than expected, and b) come up with blameless / systematic responses (e.g. internal wiki maintenance) to reduce that measure over time?