[arrowsmith]

At least 50% of debates on HN are basically: "you said that A is a good analogy for B, but you're wrong because A and B are not literally the exact same thing."

[dekhn]

And the other 50% are debates about the absolute magnitude of an effect or where a classification threshold gets set.

[arrowsmith]

There's also "you said that the average is X. But this one specific datapoint is not X!!! How do you explain that???"

I don't see that one much on HN, but it's rife on Twitter.

[cstrahan]

In this scenario, is it GP that exhibits the behavior you’re talking about, or you (given that GP gave their own counter analogy) — or both?

[_yb2s]

Sigh… when this happens I often feel like I’m done talking to people on here, but then end up coming back anyways

[indoordin0saur]

Perhaps a symptom of autism?

[allenrb]

Bah, how many autistic people can there be in one place?

[sebastiennight]

I'd say the number is at least capped at about 100 billion, but that depends on how tightly you define "person" and "place" (not even getting into the specifics of "autistic").

E.G. if you want your instances of "people" to be active, we're now capped at roughly 8 billion, since 92% of instances have already been garbage collected in this run.

I would still recommend planning a Long integer, just to get yourself some room for error.