People are the same. I believe the confusion here is thinking that the things we type and the things we say and act in person have the same meanings, both internally and socially.
Long discussion goes here, but human language is a socially-contextual performance art which has roots in the creation and teaching of the construction of tools. Vagueness, non-verbal messaging, social signaling, and all of the other aspects of language are built in order to transfer that tool knowledge with the minimum amount of effort. To be even further reductionist, you don't have to make that spear out of the exact same rock I used as long as you get a point on the end of a sufficiently long stick. Language's inability to be precise and have millions of shades of meaning based on other things besides words is a feature, not a bug. It's an evolutionary survival trait, probably needed by any sentient creatures.
Contrast that to what we're doing now on HN: typing messages back and forth. That is not the use of human language, at least not as we understand it today. Printed text is not language. Symbolic representation of sounds has a singular, yet fundamentally different goal: the unambiguous transfer of knowledge. The closer that knowledge that we want to transfer is to the human experience, the worse printed text works for that purpose. This is why fiction can be a 12-part novel while a chemistry formula might take up one page.
So when we see text messages flying back-and-forth, it engages a different part of our brain. Continuing my simplification, when we view and interact with text, we're really looking to play some advanced video game where the correct text and replies "proves" an exact version of the world that others might not share. This is why so many online arguments end up end semantic wordplay games. It's not that people are petty, it's that we're trying to take that hominid brain system and bolt it into a logical, mathematical framework. Tech folks commonly say something like "If a implies b, and b implies c, then a implies c" Unless you use a set theory definition of a,b, and c, this can be shown to fail rather easily. [1]
tl;dr, when I type the word "dog", think of a dog, and have a conversation with my grandmother about her dog, although we want to think of the word as being the exact same thing in all three contexts, it is not. It just appears that way. That's why you think people are acting so differently. They're not. It's just different things are going on in different social/physical contexts without our realizing it.
1. https://en.wikipedia.org/wiki/Syllogism#Syllogistic_fallacie...