Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Valdrax]

I have only read up until this point and I'll probably finish the rest of your presentation, though I would like you to elaborate upon your quotation.

IIRC, the number of artifacts and strange moods your fortress can get is capped by some function of the number of items your fortress has made and the number of squares excavated.  A strange mood that goes to a useless skill eats up one slot that could have gone to a more useful one. 

Even if that info is outdated / incorrect, strange moods only come about once a season at most.  I had a fortress with 3 mooded bowyers (2 of which produced [censored] blowguns) when all I wanted was to produce some good speardwarves to go to the Carnival, since crossbows are kind of pointless there.

Just be aware that if you atom-smash your, lye makers, for example, you get MORE lye-makers in the next migrant wave. Because the game notes that you have a "shortage" of that skill.

While I've never actually smashed migrants (instead favoring assigning useless ones to permanent military duty), I've found no correlation between what my fortress "needs" and what I get.  It all seems incredibly random, with lots of room for getting redundant and useless migrants.  If there was more sensitivity to that, it would be awesome.  Instead, I get a 6th bowyer for a shop which has only ever been used by moody dwarves.

One thing that would help with this is being able to request immigrant skills through the liason. That's in a another suggestion though.

It's not a bad one.  Start a thread!

I do actually think the matrix-style skill system would help things, e.g. a woodcrafter crosses over into carpentry, metalcrafter crosses over into the other metal skills, etc. Wood burner and furnace operator would cross over too.

If and only if you get the magical combination that works out for what you need.  Otherwise it's just another skill you have to grind up before getting consistent high quality.  I personally foresee a lot of migrants like glass armorers and leather carpenters.

[zwei]

Toady said that game also tries to give you migrants with skills that you use heavily.

[Neonivek]

I have a solution that is a bit easier

Migrants are given jobs and not skills.

That way the Blacksmith does both weapons and armor. The Craftsmen does bone, rock, and wood crafts. The Gemsetter prepares and fixes it.

The Cheesemaker is also a cooker, the fishermen is also a fishcleaner.

Or even that Dwarves can come with random skills or with Jobs. So you can have either an Armorsmith, Weaponsmith, or Blacksmith.

[Reelyanoob]

If and only if you get the magical combination that works out for what you need.  Otherwise it's just another skill you have to grind up before getting consistent high quality.  I personally foresee a lot of migrants like glass armorers and leather carpenters.

Now that's just being obtuse, those combinations wouldn't even exist in the RNG for a start. Unless Toady One was an idiot, and we KNOW that's not the case. A computer program can tabulate only allowed possibilities very easily. Also, rates of skill gain can be tweaked, and ALL relevant skills to a job cross-train simultaneously. There's no need to assume they're being serially trained.

But even if the game somehow did generate skill-set like that, a glass armourer is 50% to being any type of armourer, a leather carpenter only needs to buff his "wood" skill to be making wooden furniture.

And you're not considering how matrix skills systems work :-

Materials

Stone
Metal
Wood
Bone

Process

Builder
Crafter
Carver

^ 7 Skills total above, gives a max. of 12 combinations. not ALL of these have to be generated :-

Mason (Job) = Stone + Builder
Stonecrafter = Stone + Crafter
Stone Detailer = Stone + Carver

Carpenter  = Wood + Builder
Woodcrafter = Wood + Crafter
Wood Carver = Wood + Carver Not Existing in Dwarf Fortress (Yet)

Blacksmith = Metal + Builder
Metalcrafter = Metal + Crafter
Engraver(Metal) = Metal + Carver Not Existing in Dwarf Fortress (Yet)

Bone Carver = Bone + Carver
(Another 2 potential "Bone" combinations. These combinations would never actual be referenced by a job in-game.)

On "Table 2" I've listed the existing professions on the left, and the "breakdown" on the right.
That's 8 existing Dwarf Fortress jobs I've got through already, using 7 skills in Table 1, and the total skill rate gains can be tweaked to match, or exceed the current rate of gain in the game, so there's no logical reason to assume it would take longer to buff these skills. It could be instant as far as the game is concerned. There's also some "unused" combinations which no job in DF currently uses. Some make sense, some not so much.

To say every combination somehow *must* appear makes no more sense than lamenting that you might get fish-cleaner/alchemists or lye-maker/pump-operators.

Seeing as how I can create the exact same list of professions of current Dwarf Fortress on the left, that is what the game can "generate", and auto-magically give the Dwarf the "breakdown" skills needed. So no "glass armourer's" etc, that is an unecessary artifact, not part of this system.

but now there's interlinks between the jobs, both in Materials knowledge, and Process knowledge. Also, having a part-skill gives a boost to ALL related areas, eg having any "stone" skill gives you a boost to doing / learning all the stone jobs. Same with the other materials. Having any "Crafter" skill aids in "crafting" any material. A hypothetical woodcarver would get a starting advantage in both working with wood in ALL contexts but also has ability to learn carving on ANY material. So he'd be trainable to existing related jobs even if he existed but his particular skill combo was not actually implemented in the game.

So a hypothetical glass armourer would NOT be limited to "glass armour". you're not even thinking this through.
He'd be good at ANYTHING GLASS or ANYTHING ARMOUR. Any anyway, a simple job look-up table would mean you only get combinations the game actually uses.

So, effectively, matrix skill structure can mimic the EXACT same professions we have now, but with added skill interlinks. Also, there can still be jobs in the table that use only 1 skill, or 3 skills etc. So those "odd" jobs which don't fit into a "process x material" system don't need to be forced to fit, they can be exceptions or follow their own logic. Other subsets of job skills can even have their own matrix's based on any logical breakdown, or entirely left the way they are now.

e.g. Mining can still be only 1 skill even if others are using the matrix style.

[Forumite]

I like the matrix system because itīs easy to understand, unfortunately itīs complex to use. If I look at DT for a stonecrafter, then I need to check for stone and crafter, instead of just the one job. It would be nice if stonecrafters and masons had some synergy between their professions, but I think separate skills as in warfare would be better. Construction and Decorating, Stone, Wood, Bone, Metal, Cloth and Gems, where Gems include glass, or something like that.

Or we simplify the skills themselves, like in the OP, that does almost the same thing. If we still want a difference between masonry and rock crafting, then the Construction and Decorating sub-skills does that.

[Reelyanoob]

Warfare already uses "matrix style", I was considering giving it as an example in my last post, but forgot.

e.g. weapon-type(+unarmed skills) vs Fighter/dodger/striker/armor user/shield user

^ That's considerably more complex than the 2 skills per job we're discussing. I can see an armored dwarf needing to be buffed in 6 skills to be an effective warrior.

DT could be modified to give you an overall ability level too. There could be a "view by job / view by skill" option. It won't be hard for the program to give a weighted average which tells you the effective skill level. Anyway, there will overall be less skills to look at as well in a matrix thingy, so that would be less columns to go through.

[Silverionmox]

If we had three (possible) skill components (material - tool/process - product), and reduced the skills in each category to eg. a mere 10; then we would still have 1000 different possible skill combinations. This would satisfy the desire for simplicity (a limited number of skills), but still allows flexibility and specialization.

A thousand combinations?  A different skill for each material?  Good god, that is almost the opposite of what I'm proposing.  That would produce useless migrants in droves.

"Oh good.  Someone with high skills in working copper, kitchenware use, and toy-making.  On an iron-heavy map." 
"Oh hey, an animal tender who specializes in vermin and glass.  That's just what my turkey farm needs." 
"A wood armorsmith.  I wonder how he feels about enlisting in the naked shock troop squad to greet his elf friends?"

You misunderstand: in the current system, that guy would only be capable of making copper toy kitchens and a peasant for all other tasks. With composite skills, he would still be a decent toymaker, coppersmith or cook.

Secondly, I wouldn't make skills of hauling, kitchenware use, bucket use, food consumption, nose picking, marrying, etc. One can assume these are either so common or so basic everyone performs adequately: they're usually called unskilled labour for a reason. I would also put toymaking under crafting rather than it's own category, for example.

Thirdly, there is no reason at all to generate random combinations of skills. It's possible as well to generate skill with several materials, practice with several tools or knowledge of several processes.

Fourthly, woodcutters shouldn't be eager to migrate to treeless maps, just like fisherdwarves shouldn't migrate to desert maps unless they fancy a career change to sand hauler. That's a matter of immigrant selection rather than a function of the skill system. (Even so, that woodcutter would be retrained more easily into a soldier with a composite skill system because he already has skill in axe use.)

One skill for every metal is a bit excessive, that's just to show how flexible the system is for future additions and mods. You could define just one material skill, or limit it to soft organics - hard organics - mineral, for example.

Eg. Mining could be defined as stone + pick use + architecture: they'll be decent at stone jobs as well that way, and you can recruit new miners from your pool of stonecrafters, masons etc if the inevitable accidents happen. They have pick use, so they can defend themselves decently if the need arises. Architecture is to represent their knowledge of structural stability; dedicated architects would almost never need to be generated any more among immigrants, as the knowledge of the miners would suffice to handle the few cases we need architecture for right now.

One important difference with the current system is that skills and jobs will be different.

[GreatWyrmGold]

I like this idea, but maybe instead of (or in addition to?) this idea, we could have "skill synergy"--that is, a skilled metalsmith would make better goblets than a peasant, a proficient stonecrafter would make better rock tables or mechanisms than a milker or soldier without any stonework-ey skills, a potter would make better glass than a clothier (maybe, I don't know much about how RL-pottery or glass industry works), and so forth.
Depending on how far you wanted to go, skill in carpentry might slightly improve one's ability to make stone items that can be made out of wood, and so forth.
-----
As to the original idea, I really would like it if the farming labors weren't so fine-grained and if the fish cleaning/dissection labors weren't separate (which one un-raws fish again?), and so forth, but wood/bone/stone crafting should probably be kept at least at arm's length, so to speak, although making bolts/arrows/etc should probably be moved to bowyery.

[Reelyanoob]

Skill synergy by itself, or a multi-component skill system we're discussing here, can both be set up o achieve the exact same result / outcome.

The main difference is that with "skill synergy" where you have skill level bleed over to new skills is that when you add a new skill you have to make a matrix which describes how much it bleeds over into all other existing skill (so you end up with a vast list of "related" skills for each other skill to remember)

Also with 'skill synergy' it suffers from a lot of issues / questions which the skill matrix idea does not.

You have to as what's the formula for the synergy? There's lot of possibilities and contradictions. If someone is a novice mason but level 10 in engraver or stoneworker, you could say they start as "effective" level 2 (1/5th) in Mason, and say the same thing for Stonecrafter. But what if they're level 10 Engraver AND Stonecrafter? Do you start them as effective level 4 in Mason then, or cap it at the highest synergy level, e.g. 2? And what happens as they gain +1 level in Mason, do you add this to the synergy levels or not? e.g. they would be effective level 5? Or is that effective level 3? That means you'd HAVE to cross train the other related skills to max out the one you want. Unless you cap it to the highest of synergy levels and REAL levels, but that would bring in it's own weirdness. And if you add a new 'stone' skill it has to be explicitly X-referenced with all other related skills. If the synergy amount is cumulative, then effectively you can cheat by making 40 different stone-related skills and get a higher level in Mason, which you haven't even trained at all. The mere existence of new skills throws the synergy system into confusion.

^ The above is a synopsis of the main issues with a Skill Synergy system, you have to do the math & consider how such a system will work in the real game.

The skill matrix however, does away with these problems. Because each skill is made up of shared components. And you've pre-trained the components themselves, so effectively you're just saying that the character starts out with some extra relevant skills, he's not a skill level 0 mason who magically can do better because he made some stone goblets, while his character info says "zero experience" as a mason. And as he trains up the other component of his skill he will gain ability in a smooth fashion.

 So you get the cross-training without the confusion, and you don't need to scan the entire list of skills each time you do a skill check. (you'd have to scan the entire list for crossover skills or otherwise hard-code the lookup tables, and that would be worse)

[Forumite]

I like giving a x2-x5 bonus experience (depending on the difference in skill) when training a related skill that is at a lower level, so the Legendary Mason can train the first levels of  Stonecrafting 5 times as fast as the latest soapmaker immigrant, and will on average become legendary in a third of the time.

[Wimopy]

I like the idea in general, although I do disapprove some of it, which I will list in an edit tomorrow, too tired for it now (I hope I remember).

What I was also thinking though, that while we make these changes, we shouldn't necessarily merge them fully. What I mean is this:

Make it so, for example, crafting wooden bolts also gives some minor metalcrafting XP (dunno if someone said this already, as I said I'm tired) along with some other crafting XP, since you don't only learn how to carve bone, but also to shape things in general.

That's all for now, hoping to edit tomorrow.

[zwei]

I like

I like the idea in general, although I do disapprove some of it, which I will list in an edit tomorrow, too tired for it now (I hope I remember).

What I was also thinking though, that while we make these changes, we shouldn't necessarily merge them fully. What I mean is this:

Make it so, for example, crafting wooden bolts also gives some minor metalcrafting XP (dunno if someone said this already, as I said I'm tired) along with some other crafting XP, since you don't only learn how to carve bone, but also to shape things in general.

That's all for now, hoping to edit tomorrow.

This is problematic in the issue that bleeding skill xp vaguely related tasks open weird can of worms (if i craft wooden bolts, what does that teach me that i can apply to crafting metal bowls? nothing, really...).

Not tm mention  that it is quite hard t communicate to player what happens (huh, why is my legendary woodcrafter that only ever made bolts also master metalcrafter? is there a bug or something?).

Skill matrix idea, whic i started to really like makes it easy to grasp reasons why skilled woodcrafter can also be passable metalcrafter ... or why he is not if skill matrix happens to be completelly separate.

Granting XP to relevant skills is also easily understandable.

And if each matrix skill is internally implemented by ~32 bits of "skill bit" system, all the better.

Besides, matrix skill system solves problem of new arrivals - something that synergies do not. Example of one skill matrix in use - combat skills is quite persuasive. Even if immigrant arrives as skilled blowgunner, he still has enough skills to be great choice for crossbowdwarf and is decent choice to be melee fighter.

I can't wait the day when butcher arrives, but i have too many butchers already to i can make him cook isntead (since he knows how to handle meat) or maybe even bonecarver (since he also knows a bit about bones).

[Aquillion]

Most of these are really good ideas, so I've only focused on the one I object to (since there's not much to say about the ones I agree with!)

My main feeling is that differences between skills should be eliminated when they don't represent a real "choice" on the player's part -- fish cleaning and fish dissecting, say, should obviously be merged, because you're just going to train them both any time you want either, reducing them to pointless fiddling and tracking on the player's part rather than a real choice.

Whereas, say, the choice between designing your fortress to make clothes vs. designing it to make leather armor is a major decision, and should be preserved.

Proposal 7.  Merge Leatherworking into Clothesmaking?  (Retain separate shops.)

No.  As I said above, these skills do very different things, and represent a substantial difference in focus for your fort, both in terms of the materials you're building your economy around and in the goals your economy is focused on -- making players make a 'guns vs. butter' choice between them is, I think, a good thing.

[Valdrax]

If and only if you get the magical combination that works out for what you need.  Otherwise it's just another skill you have to grind up before getting consistent high quality.  I personally foresee a lot of migrants like glass armorers and leather carpenters.

Now that's just being obtuse, those combinations wouldn't even exist in the RNG for a start. Unless Toady One was an idiot, and we KNOW that's not the case. A computer program can tabulate only allowed possibilities very easily.

Just because none of those combinations are usable together doesn't mean that you can't have a multi-talented, jack-of-all-trades, master-of-none dwarf.  After all you can have under the current system an Talented Carpenter who is also a Novice Brewer, Weaponsmith, Mechanic, and Glazer.  Or a High Master Milker and Competent Fish Cleaner.

Those are two of my latest migrants.  There's no synergy between their skills right now, without multiplying the places at which the RNG can screw you over, which is what your proposal suggests.  Having actual grouping of skills in migrants beyond "a little bit of everything under one job category," like the "Novice Omnifarmers" you get sometimes would be nice too, but the game as stands does nothing to prevent random, nonsensical skillsets.

But even if the game somehow did generate skill-set like that, a glass armourer is 50% to being any type of armourer, a leather carpenter only needs to buff his "wood" skill to be making wooden furniture.

Or to put it another way, said armorer has half the experience that one who didn't need a separate materials skills would have towards making quality goods.  Rather than either being good at making a useful product or not, you've got more hoops to jump through.

Simply combining skills and ignoring materials handles most of the same issues.

No, I get the idea of having a matrix of skills leveled at the same time for synergy.  I'm not disagreeing with you because I don't get the idea.  I just think...

• It adds interface confusion and complexity, partially thanks to incomplete material/job grids (Leather Furniture?) and partially thanks to one-material skills, like Mechanics.
• It makes some skills less useful than they already are.  (e.g. Glassmaking, Leatherworking.)
• It focuses mostly on breaking up skills which produce items with quality, complicating leveling them.
• It fails to address most low-utility, quality-less skills like the excessively fine-grained farming skills.

That last one is important to me.  You completely ignore the quagmire of useless skills for farmers, rangers, and fishers (e.g. the widely hated Fish Dissector.)  The material system does little to fix that, and proposing another incompatible, independent matrix of food skills on top of that would only produce more confusion.

[Valdrax]

My main feeling is that differences between skills should be eliminated when they don't represent a real "choice" on the player's part -- fish cleaning and fish dissecting, say, should obviously be merged, because you're just going to train them both any time you want either, reducing them to pointless fiddling and tracking on the player's part rather than a real choice.

I think that's a pretty good summary! 

Proposal 7.  Merge Leatherworking into Clothesmaking?  (Retain separate shops.)

No.  As I said above, these skills do very different things, and represent a substantial difference in focus for your fort, both in terms of the materials you're building your economy around and in the goals your economy is focused on -- making players make a 'guns vs. butter' choice between them is, I think, a good thing.

Yeah, I put a question mark there because I was iffy on it -- more because of the differences in actual tools and techniques used.  However, you raise a pretty good point there against that proposal as well.  I might delete it eventually if there's enough opinion voiced against it.