As you all know, the current LP system mostly sucks, and along with it the systems of buying skills and thereby of gaining crafting recipes, combat attacks and constructions (collectively: "actions"). The crafting system itself also has a quite a few aspects that are less than optimal. Some of the specific aspects that I'd like to fix is that (while it's fun to be rewarded for pretty much anything) gaining LP is mostly grindy and therefore very macro-friendly, and the fact that it isn't meaningful (or even possible) to specialize in any way. In particular, if it were meaningful to specialize a character, that would in turn make it meaningful to replay the game from scratch from time to time (or from not-entirely-scratch when dying).
Our current thoughts in rectifying the situation circulates mainly around two areas.
First of all, we should introduce deeper character psychology. One psychological aspect we're considering is "work weariness" (or maybe "world weariness") vs. "fun", which would be the foundation for replacing LP. Work weariness is gained when doing stuff, basically; kind of like stamina drain, but differently proportioned. You don't want to be to weary of work, because that is bad; though it is yet unclear just how it is bad. Maybe the character will refuse to do things; maybe he'll get epic debuffs; maybe he'll spontaneously suicide. Work weariness would be reduced by having fun. Fun can be had in various ways, from eating good food to playing with toys to killing people. Reasonably, the personal beliefs should determine what causes work weariness and what kinds of fun the character prefers. When restoring weariness by having fun in the right ways, the character should gain some substance of thought -- let's call it "Spiritus" just to have a term for it -- not entirely unlike LP, but just how it should work and for what it should be usable is a bit unclear. It should be usable for increasing stats, but it may also be usable for various other things, and it may also have some finer structure so that various forms of Spiritus is better used for some stats or things than for others.
Second, actions should be learned individually and not come en masse along with a skill. Maybe the current skill tree will disappear entirely, maybe it will have other uses (we have some ideas); that is not yet clear. Either way, learning an action should be more involved than learning skills currently is. As described briefly here, they will probably be learned via another part of character psychology currently going under the working-name "curiosities" -- a quest-like system. To quote myself from the link:
loftar wrote:[...]the kinds of quests we've been considering so far are not NPC-based. Rather, we've been thinking more and more of introducing psychological aspects of characters [...], and that these kinds of quests would probably show up as things that the character has become "curious" about. Like "Hmm, this 'clay' looks interesting. If I dig up a couple of more pieces of clay, I might think of something to do with it."
Individual actions will probably also be endowed with per-action quality, that are determined when learning the actions. It may be possible to increase the quality of actions somehow, probably through other curiosities and limited by some character stat(s). In particular, since it should be harder to learn individual actions, it should be possible for characters to teach actions to other players, which will transfer them with lower quality.
I hope it is unnecessary for me to point out that an important point we will keep into consideration is the fact that it must be fun, as opposed to tedious, to learn new actions, and that the game experience must remain meaningful even without a lot of individual actions.
In relation to the revamp of actions, we have considered some revamps of the crafting system, though much is yet unclear about it. The high-level, abstract idea is that it would be fun to be able to learn "master recipes" every once in a while. One such master recipe might be "bread". Such a master recipe would take generic inputs such as "flour" (assuming there are different kinds of flour, such as wheat flour or barley flour), "liquid" (it may be possible to choose between water or milk, for instance) and "flavors" (where one may be able to put in stuff like butter, poppy seeds, honey, and what not). Generally, each combination should yield, not a bread, but a specific recipe for that kind of bread, where the product would have different qualities depending on the inputs. It should probably be a rather involved process, somehow, to create new specific recipes. It should probably be possible to name the specific recipes, so that I can call the bread I've invented "Loftar's Lembas" for real this time, and then teach that recipe to others under that name (and also have it reflected in the tooltip of the product). Certain very specific inputs may yield not new parameterized bread recipes, but specific recipes that we've pre-seeded, so that wheat flour, water, butter and apples would result in the recipe for apple pie. That would make the game more rich in exploration and experimentation. The specifics of this scheme are yet unclear, but I think it should be possible to bring it to concretes, and I think it would be very cool indeed.
Another point that is related to learning actions individually is that the available curiosities should depend on the resources available in the area. We plan to combine that with a new map generator, which would generate very large "terrain zones" or "climate zones", if you will, where each of the various terrain zones would contain different base resources. These would not be entirely different resources, but rather differences in terms of alluvial clay vs. varve vs. quick clay; or firs vs. pines vs. junipers. The various resources would result in different actions being learnable, and while the kilns makable from one type of clay may not be fundamentally different from those makable from another type of clay, there should still be slight differences.
The variations in base resources and actions learnable from character to character should be able to result in the cultural variations alluded to here, and should give rise to possibilities for character specialization.
As I've mentioned in the above linked thread, however, I am very fond of the current system of character creation where you choose virtually nothing at all -- just a name and a gender, where the gender is purely cosmetic anyway, and the name is hardly visible to anyone but yourself now with the new, nice Kin System. I like it very much in that one makes no binding choices at all (that matter) during character creation, and that all choices come as part of players' interaction with the world instead.
Last, it is worth mentioning that one thing that is probably not a problem with the LP/Skill system is the fact that the early game is more grindy than fun and consists mainly of lighting fires and building wicker baskets to gain some real skills. Rather, that is more likely a problem with there being too few things to do, quite simply. As you've seen, we've started to remedy that lately with more noob stuff (not to mention wilderness spawning), and I think we can all agree that the early game is a lot more fun now than it has ever been before.
