The first (and last) time the rules defined hidden as a "condition", or explicitly stated that if you were not hidden people knew your exact position was 4E.
In context it wasn't quite that unconditional (ie everyone on Oerth doesn't know the position of everyone else who hasn't made a stealth check recently, all the time). That shouldn't have to be stated, but this is rapidly becoming one of /those/ threads.
But, yes, Hidden was a condition, with an explicit game-jargon meaning, and, after some confusion very similar to what we've had with 5e this time around (afterall those specific rules didn't change that much 4e->Next->5e), there was an errata that clarified that meaning (or maybe even established it, I don't recall atm, it's been settled for so long). After that, the issue was no longer in question.
That's one of the benefits of a 'gamist' design that defines and consistently uses (and promptly errata's when there are issues) a bunch of jargon. 5e didn't go that way, it chose natural language over jargon as much as possible. That makes the book a more accessible read, but a more ambiguous source of rules. In turn 5e leverages that ambiguity as part of the foundation for DM Empowerment. You /need/ the DM to make rulings where the rules are ambiguous, that dependence empowers the DM to take a more active hand in shaping the game on all levels.
When folks trot out claims of 'RAW' they're - probably unintentionally, perhaps out of habit - actively undermining that goal of DM Empowerment.
OTOH, while Hidden is not a condition in 5e, Hide /is/ an Action. So there is an explicit meaning, there. FWIW. (In 5e, not much until the DM makes his ruling!)
So I make the assumption that people state that your position is automatically known unless you've taken the hide action because they got used to 4E. No one has pointed to any rules in 5E that explicitly state otherwise.
No rule explicitly states that DMs who rule one way do not do so out of habit formed in an prior edition? I mean, AFAICT, that's how that last paragraph parses. And that last sentence is true. An odd thing to say, perhaps, but true.
Well, that's one way to parse it - I'm the one who's going on about the ambiguity of natural language, afterall, I should allow that there are others....
Another way that also fits would be that "no rule in 5e explicitly contradicts the assumption that your position is automatically known unless you've 'taken the hide action' successfully." That's also true.
I'm guessing you meant the opposite, that "no rule in 5e explicitly contradicts the assumption that your position is not automatically known even if you have not taken the Hide Action." Which is particularly true, I think, given the trivial case of obviously not knowing where every non-hiding creature in the world is, all the time.
More to the point, it's up to the DM to rule whether the position of a creature is known to another or not - whether someone has declared a Hide action or declared searching for a hidden creature as their action. That ruling can include calling for check from either or both parties. Or not.
(Though I'd advise against call for checks 'both,' as I consider opposed checks overly random/'swingy,' especially under BA. JMHO from years of experience with 3e.)
As to the a question of whether a second character knows a first character's position after the first casts invisibility, I think that needs to answered carefully, with a difference between "knows definitely where a character is located" and "has a good idea of where a character is located".
In another thread I brought up the possibility of knowing that the character is still present, but not the 'exact' location. So "you hear footsteps echoing around the cavern, you can't tell where they're coming from..." or "... you're pretty sure he's in /that/ side of the room somewhere..." Facilitates recon by fireball, that.
For example, if there was a higher level "invisibility and dimension door" spell which made a character invisible then moved the character 10' in a chosen direction, would a character automatically know the final location of the character who cast the spell?
There were powers like that in 4e, and if they were meant to prevent others from knowing the final location of the user they would contain language like "...and you can make a stealth check to become Hidden" or just "...and you become Hidden until..." That's jargon and exception-based design for you.
Something similar in 5e probably wouldn't contain such language, and leave it to the DM to rule how the teleportation & invisibility interact. For instance, one DM may decide that teleportation opens a dimensional gate and/or bleeds off interdimensional friction as a puff of smoke and/or displaces air with a loud pop or whatever - and thus the two points involved are obvious, even if the character teleporting isn't; another DM might rule that the obvious intent of the spell is to escape detection, so it must render the caster's location unknowable short of more powerful magic.
That's natural language and DM Empowerment. The very existence of a spell like that, worded that way (in ambiguous natural language) presents another opportunity for the DM to fine-tune the challenge of an encounter, the tone of the setting, and/or the nature of magic in his campaign.
