But in dnd (or any other pnp game), your fate should be determined by the most fickle of women (ie: lady luck), in the form of dice. You can't say "I am meant to be a hero
I honestly am flabbergasted (That's a fun word). I can understand completely that other games have no deaths, in the example of the Teenagers thingie. Seriously, I back that 100%, if the game isn't meant to have deaths, go for it.
But D&D IS meant to have deaths.
What is the basis for claims like "should" or "can't" or "meant to". These look like nothing more than statements of personal preference.
That why we have raise dead spells and rituals, and rules for when you're unconsious and when you die.
Well, as Mustrum Ridcully noted upthread, readily available raise dead spells are functionally a type of (potentially verisimilitude threatening) death flag mechanic.
The desire to continue experiencing the campaign's story using the same character is no more important than the the desire to continue experiencing the campaign's story while not being lost in the dungeon, eating rats, and being chased by a minotaur.
What is the basis for this judgement of relative importance? Presumably a player playing in a death-flag-style game of D&D has already decided (i) that they want to use a particular PC as a focus for participation in the game, and (ii) that they do not object to the GM throwing in complications such as being lost in a dungeon eating rats and being chased by a minotaur.
Which begs the question of...are these "consequences" for failure in any way meaningful to the player?
If they are not, then (for reasons other posters upthread have already noted) the game will probably not play in a very satisfactory way.
If we say, X is a consequence of failure, and you say "I don't like X; let's just have Y", why would it not be equally valid to say "I don't like Y; let's have A" (where each letter is progressively less of a consequence as one descends from Z to A)?
Maybe it would be. But death-flag play is not motivated by the thought "I don't like Z; let's just have Y," in which "Y" is Z-lite. Death-flag play is motivated by the though "I don't like Z, so let's have Y instead" where Z is something disliked (eg thematically/aesthetically unsatisfying play) and Y is something like (eg thematically satisfying play, in which thematically arbitrary PC death should play no part).
What is the qualitative difference between them? None. Why is "giving players authorial control w/r/t character death" the cut-off line?
Typically it is not, although for some players it may be the most important issue - because if the player is allowed to keep the same character, AND if the player is allowed to develop the backstory and the ongoing story of that character, THEN the player automatically enjoys quite a bit of further authorial control. But to see further examples of player authorial control in death-flag-style play, consider the discussion of "fact introduction" in the Challenge the Players thread, for example. Or look at the "raising the stakes" rules which, together with the death-flag mechanic, are part of E6.
Unless you never kill anything else, I don't see why they'd hold back from killing you.
Who said the monsters are holding back. They're not. It's just that the mechanics favour the PCs. A pure deathflag mechanic (or Fate Points or Karma Points or whatever) is a metagame mechanic that favours the PCs. 4e D&D uses a bunch of mechanics that mix game and metagame (hit points, healing surges, saving throws, and all the powers that interact with them) to produce something that is closer to the death flag end than the low-level AD&D end of the spectrum.