And my point was that immutability isn’t granting any performance here. It is, in fact, causing a performance detriment for no reason other than “a medium dudebro wrote an article that included the words beautiful and elegant, and our engineers are idiots that eats that brain dead nonsense up”.
The whole article is, hilariously, about how shitty immutability is, so they hack and slash and bend over backwards to make idiotic nonsense comport to their needs, when had they just fired all their medium reading seniors off the bat, they wouldn’t be in a position of spinning up 10x the CPUs to mangle their data around this bullshit.
> -read speed on aligned, flattened data
This is the case outlined in TFA.
> But, that ignores that fact that getting to a point of aligned, flattened data with immutable structures is incredibly slow.
It might, if that weren't the whole point of TFA.