thanks, i don't like to argue anyway
Cool. I'm not a huge fan of it either because almost invariably it ends up being meaningless.
Well, i think there should be SOME class specific abilities for every class...
Agreed. If everyone can do everything, then you might as well have a point buy system and bring along all the baggage and problems that come with it.
But may I suggest to general guidelines:
1) Class specific abilities should be limited to things which differ in quality, and not merely in quanity. For example, swinging a club ought not to be a class specific ability. Rather, how well you can swing a club can vary between classes. If you find yourself defining as a power something that is mundane, you tend to run into problems where players are forced to interface with the game world purely through meta langauge. That is to say, rather than being able to say, "I do this in the game world.", they have to say, "I activate my power." And likewise, rather than the DM being in the role of interpreting the players proposition into mechanics, the DM merely ends up as an accountant that tracks the results of mechanics.
2) While its perfectly fine to have independent subsystems for handling things that are different, there should be no more than one subsystem per in game reality. "There is more than one way to do it" leads to confusion, rules bloat, and imbalance.
From those guidelines, its easy to see where you have silo and where you have overlap.
Take the 4e Fighter ability Combat Mobility (i think that the one, maybe i have the name wrong). The fighter could stop someone from moving past him with an OA. No one else in the game had that ability, it was the fighter's "thing".
Excellent example. The question is, why should a mundane thing like "I interpose myself between X and the charging Orc" be an something exclusive to a fighter? I mean, I would think that even a 5 year old could attempt such a manuever, though granted they probably wouldn't be successful unless the attacker was another 5 year old.
See to me, in order to really have a shot at being considerd a Role Playing Game and not merely a tactical board game that allows for role play, the game has to procede from proposition to rules without the rules getting in the way of reasonable propositions, rather than from game rules to player propositions. If the only thing that the players are allowed to try are the things on a list of powers, then you don't have a game which at its heart is an RPG.
So in my opinion, if you want something like an OA to potentially stop movement at the point it occurs, then you need a more generic mechanic for that (or at least more generic access to the mechanic) and maybe the Fighter's schtick is that he can be in some way better at it. Maybe you have a generic combat manuever check and the fighter gets a bonus on all combat manuevers. Maybe you have a Tactics skill that as a skill check allows you to stop movement with a successful OA, and the fighter gets unrestricted access to that skill (a 'class skill') while no other class does. Or maybe the fighter gets to select a combat bonus or feat every level so that he becomes extremely competent in the arenas of combat that he chooses, and no other class can match that degree of prowess.
All of those work better than making a generic combat action something that can only be done if you have a particular power.
As an aside, one of the things that lets you figure out that a powers system is simply a magic system in disguise is precisely that instead of making you better at mundane actions, it provides for performing actions that are impossible for characters that lack whatever esoteric skill, knowledge, or ability lets the character perform the action. An example would be allowing the character to throw a grappled opponent, something otherwise not provided for in the rules. Under such rules, 'throw opponent' is basically identical to a spell, and the game will play as such, ei: "I cast my 'throw opponent' spell"/"I use my 'throw opponent' power.". Another clue is that the rules provide no chance for the power to fail situationally. An example would be a power that lets you quickly shoot arrows into the wall to make a ladder you can climb, with no mention whatsoever of the hardness of the material you are shooting at and no attempt to reconcile the fact that on a normal round you can't in fact shoot 12 arrows in the same round with that sort of precision and force nor any allowance of letting you use this flurry of shots for any purpose but ladder construction. This approach is basically summed up as, "Well, balancing spellcasters with nonspellcasters is hard, so rather than tackle the problem I'm instead going to simply make everyone a spellcaster.", and Mearls is one of its chief practicioners.
Finally, the overriding clue that you're mundanely dressed powers are actually spells in disguise is that there metagame restrictions on your access to them. This usually comes in one of two forms: either a time based restriction on how often you can use the power which is based primarily on how powerful or useful the power is, or some sort of token, slot or points which you have in limited quantity and which you must spend in order to activate the power. (These points usually renew over time but sometimes in exchange for some other game resource).
Now, I'm not saying that 'Everyone is a spellcaster' is all bad. It has some advantages; but, I am saying that I think 'everyone is a spellcaster' and its associated baggage is one of the more important reasons why a large number of people who tried 4e said, "This isn't D&D" or even "This isn't an RPG", and if you go that route you won't be able to pull those players back into the game.
