Choose Boring Technology and Innovative Practices

I think on the «materials» side there is also an additional question: how churn-y is the thing you consider trying.

Old Julia stable versions go EOL relatively quickly, system update compatibility will catch up with you, and next versions require you to update your code. But if noone knows how to change a weird FreePascal program anymore, well, you stop changing it — but update the compiler as needed, and chances are very good the program will work as before.

This even somewhat cuts into practices — if operating the system is tied too closely to some practices, just abandoning the practices might not be a done-in-a-second thing.

I certainly don't disagree. The "tools" side is the place where I have had the most freedom including trying things not used in the company otherwise like Golang and Flutter. It's not even the inherent risk on the tech alone. It is a lot harder to find a competent Scala developer than a C# developer which adds additional risk outside of EOL and lack of long term support causes.