But they don't have to tell us what a safe place is in order to be built around the idea of "to regain hit points you need to be resting in a safe place".I don't really like the rules defining what does or does not constitute a "safe place" outside of specific adventures or campaigns.
For the past several five or so levels the PCs in my 4e game have been in the Underdark. They can only take an extended rest in a "safe place". What sort of place counts as safe is flexible, and has been worked out in play: so far a Hallowed Temple (created by ritual), a duergar stronghold and a drow barracks have all been used.
I am very much hoping that the game won't be built around an assumption that recovery rules for monsters - to the extent that they are needed at all - are the same as recovery rules for PCs. That's an optional approach to play, obviously, but it should't be built in as a rigid assumption. And those who are using it will probably not be using a "havens" rule - they'll use something more simulationist, perhaps like [MENTION=882]Chris_Nightwing[/MENTION]'s idea upthread.The main problem I see is what about monster HP. A group of orcs take an ancient fortress out in the wilderness. They are not in posh comfortable living conditions so now they have 1/2 HP. In fact most monsters have 1/2 HP now, only those that must be repaired or regenerate is some way or living in comfortable civilized conditions get full HP.
In 4e the opposite is definitely true - that dealing with 4 ogres together is harder than dealing with four lots of one ogre - because the ability to optimise area damage is not as important as the ability to optimise action economy.an encounter with 4 ogres is probably easier than 4 encounters each with 1 ogre (because the wizard can usually affect more than one with his fireball but probably doesn't have four fireballs to throw around).
If D&Dnext is trending the other way - that wizard AoE damage is more important than overall party action economy - then it's moving away from my own preferred destination.