[silvarilon]

Exactly. I'm a programmer, so I can certainly code in more behaviours. But I need to think of them, first. (Hey! I'm a "logic first, creativity second" kind of guy!)

I'm sure we can come up with more behaviours. Something I'm planning is "fashion" - each item of clothing would have three stats. What rank it's intended for, how fashionable it is, and what colour it is. Then each month the "fashions change" to make different garments more or less fashionable, different colours, etc. - so the "quest" would be for the players to figure out how the fashions changed. Maybe doublets are still in vogue, but red is totally out, and blue is in.
Then there would be clues. The NPCs might react when someone enters the room if they are wearing something particularly fashionable or unfashionable. John walks in wearing a blue scarf "Oh dah-ling! I lurve your scarf!" - you turn up with your red doublet "My my my, that doublet is just *so* last month."

Essentially I suppose that would be a "gather item" quest with a puzzle to find what item to gather.

Oh yes. We certainly do human-overseen quests all the time (fully 1/3 of our staff is dedicated to that. And players can submit their own story ideas that they want to play through and staff try to support those.)

I tend to think in terms of "stories vs quests" - in my mind, a story is an event that happens to a character. That event might be "I found the boy's lost dog" or it might be "I met this woman, we talked long into the night, and I fell in love" or even just "I got in an argument with someone in the street today"

To me, a quest is something coded (or prepared by a human) to give players an activity to do. That activity could then lead to the story. So the quest would be "a boy asks you to find his lost dog" or "the crazy drunkard tells you to run through the streets naked" - hopefully the players then take those quests and turn them into stories.

A quest with good story structure will, of course, make for a more satisfying story for the player.

Oh, of course. Depending on how your game is set up, you want to focus on the core gameplay. Quests resetting is very much a minor topic in comparison.

Heh! Sounds totally awesome.
At the risk of derailing the conversation, could you go into some detail about what the players do to solve the mystery? Do they have to collect certain items (clues) - or are there actual clues hidden around that the player has to figure out themselves? How does the system know when the quest is completed?

Yes, but that's also because of the type of game. If you're playing WoW, your "core gameplay mechanic" is to go out there and kill monsters. NOTHING should get in the way of the players ability to go out, find monsters, and attempt to kill them.
A lot of MUDs are set up similarly, your core activity is to kill things. Fight your way to the boss, and kill the boss.
But not all. My mud is extremely social. The activity of killing the boss isn't nearly as important as the impact that will have on the rest of the city. Does your character get lauded by all the other characters for their bravery? Was the boss truly evil? If you kill the Goblin King does that help the city or hurt it? (Are a whole lot of goblins fighting for leadership worse than one coordinated group? Maybe the goblin king kept the goblins into the forest, and now with all the fighting amongst goblins they're running low on food and raiding farms?) - so in my MUD players would mind very much if the goblin king returned.

That doesn't mean the quest can't reset, though. It just means I'd plan it to have consequences, and I'd make the quest a more long-term thing.
For example, you do the "kill goblin king" quest (large fight). You win. Dead goblin king.
The "kill goblin king" quest doesn't reset automatically. Instead it starts the "save farmer's daughter from goblins" quest (simple fight), the "hungry woodsman because goblins fighting are chasing off all the game" quest (gather food), the "defend the city" quest (simple fight, PCs sent out to drive the goblins back), and the "help goblin gain power" quest (the goblin might ask for you to give him weapons and food, so he can gather a goblin army and become the next goblin king. difficult gather items quest)
The more players complete the "defend the city" quest, the less goblins there will be hanging around, until goblins stop attacking the farmer's daughter (and that quest vanishes) - but if someone completes the "help goblin gain power" quest, a new goblin king will rise (and that player will be richly rewarded by the new king) - and the quest chain starts again. We can randomly generate "goblin names" so "Grozzold the Goblin King" might be dead, but now "Ziggit the Goblin King" is in power. Silly, short-lived goblins! More importantly, to me, we've given the possibility for social stories. How do other characters react to the PC that helped a new goblin king rise? Do they flock to him (with his newfound NPC ally, he might be the only character that can provide goblin poison arrows, or something...) Or do they despise him? How will he react if Ziggit the Goblin King, who provides him with poison arrows, is killed by another PC? If he tries to raise the next goblin king to power, will the blacksmith's refuse to sell him the weapons he needs?

But... you can see the difference in emphasis. In my example, the emphasis isn't on killing the monster, it's on creating social impacts from player actions. Because I'm running a social game.
In a fighting-oriented game, you could just have the quest reset, with a new goblin name, and say "A new goblin has fought his way to the top of the pack"

Yup. I'm quite a fan of persistent quests. When I play a game, I like feeling that I had an impact on the game world.
It frustrates me if I'm playing a standard game where my fighter goes out, kills *every single goblin* in the forest, and tomorrow the forest is still equally full of goblins. Makes me feel like my fighter may as well have just stayed home. On the other hand, if my fighter goes out, kills goblins, and recaptures the lookout tower from the goblins... and tomorrow when I go out, there's a human sentry in the tower instead of a goblin sentry? That's satisfying to me. It feels like I made a difference. And if I stop killing goblins, and they take back the lookout tower? That's also fun. It makes me feel like I was useful and needed, and the world is worse off without me helping.

So a plague of undead that stops when the priestess is slain? Awesome! A new priestess might turn up at some point, sure. But until that happens, the game world is different.

Hard to do in huge games like WoW but fortunately (unfortunately) most MUDs have significantly smaller player bases.

Two questions:

1) When you guys make quests, do you expect that each character will do each quest once? I.e. arrive in town, find lost dog, fight off undead, save farm, go into forest, kill goblin king? - or do you assume that there is only one "quest" of each type at any given moment? So if two people arrive in town, the boy wants his dog back, but only one of those two people can complete the quest?

2) How do you handle puzzles? For example, say I have a puzzle that requires me crossing the river by borrowing a boat from a farmer, but the farmer will only give me the boat if he's in a good mood, and I need to bring him a beer to get him in a good mood (with the pile of empty beer mugs outside his house being my clue) - What stops players just telling each other what to do? Is that even a problem, or do we just assume the player misses out on the fun of the puzzle, so it's their own loss? Do you use any code or randomization to try to make the puzzles different to stop players knowing the answers without figuring it out? (such as my above example of fashion where they can know what the puzzle is, but still need to figure out what the current fashion is.)