Smithing & Anvils alternate

[redram]

I wanted to suggest an alternate system for smithing, and I figured I'd best do it now while survival is the focus. I must admit first that I am not a coder, and also I have not played vintagecraft. Only vintagestory. My VS smithing knowledge comes only from watching VC videos. That said, hopefully I can still contribute some useful thoughts.

THE PROBLEM
 

Spoiler

 

The problem as I see it right now, is that the VC smithing system (to which VS aspires right now if I understand correctly) – is basically a stripped down version of TFC. The TFC system is a fine balancing act: randomized target, keeping the forge fueled, keeping the ingots heated correctly without melting, keeping flux in play, and a strong hammer. It's something of an art irl. The VC system removes the discovery element of a randomized target, and removes the tension of melting ingots. It removes flux. You're basically left with a system of heating the ingots, making sure you've got a strong hammer, and playing Simon Says with some buttons. It's not very compelling. The buttons use the same symbols as TFC, but have lost all the meaning of moving a target by different amounts, in different directions.

Now, my impression of VC/VS right now is that it attempts to remove some of the frustrating aspects of TFC, of which smithing can be one. Specifically the randomized target. Fair enough. The interesting thing about this though, is that it opens up an avenue for a smithing interface which is more 'real', which TFC cannot do if it is to maintain it's random target element. I think my proposal would help to make the smithing system more engaging, but without adding frustrating randomness.

THE BASIC PROPOSAL

The basic proposal, is to make the interface simulate moving metal around, to arrive in the shape of the desired product. The player would use various tools and/or moves, to make the metal do different things. This may occur in a GUI, or it may occur in a more 'real-world' environment, the way that some minecraft mods overlay GUI elements in the character's normal view. I'm going to show some crude graphics to help illustrate what I'm suggesting, and they will basically look like GUI stuff. But – and here is my non-coder-ness showing – they may be able to be translated to a GUI-less format.

DETAILED EXPLANATION (extremely long - sorry)

Spoiler

 

So to start with, you'll have the anvil top, and a hot ingot. Using a 16x16 work area, you might come up with something like this:

Now following is where you need the orange to be.  A 'recipe' if you will (in this case, an axe recipe).  The yellow shows the area you need the orange to be expanded to.  This would not actually show in the interface.  The player remembering it is part of the trick (or, maybe low tier anvils will show recipe outlines via menu selection).

Now the commands.  First two might be Hit and Upset.   Hit, when applied to an orange square, creates 8 orange squares around it.  This is great for quick expansion.  Upset creates just one more orange square, in the square directly 'up' of where it is applied.  So the first several moves would be hits, applied like so:

But we see that the last hit was too much.  It created orange where we didn't want it to be.  Now a third command, Split.  This command removes a square of orange. 

That puts us back where we need to be, and now we start using Upset to expand into the smaller areas (if we'd not used the erroneous hit, there'd be three upsets to finish the top tip):

But now lets add another rule.  Upset can only add an orange square, if there is not already another beyond it.  This means we can't use upset on the remaining four yellow square currently, because we already had the rule it only moves the 'up'.  This rule means we don't have to have four upsets for different directions.  So instead, we add arrows to allow us to turn the work piece, and we turn it to the right, and now we can use upset to fill in the notch.

Now we turn it again so the remaining three yellows are 'up' from the piece.  But let's add another rule: all strikes must occur over the grey area of the anvil.  Now we also need to be able to move the piece itself up and down.  So we add those buttons, do the remaining three upsets, and we've completed the pattern!

Now, we could stop here.  That's a fair amount of work, definitely more than the current system.  But, not random, and should not be as frustrating as the tfc mechanic.  At the same time, it provides a much higher degree of similarity to how actual blacksmithing works.  But...we don't have to stop there.  We can now add finishing moves, shown in yellow on the left (and also a fourth basic move, punch, which is used to create holes - not used in the axe) So now we have to finish the sharpened edge of the axe.  We use the DRAW command to do this.  It's many strikes:

Now we also use BEND, to bend the back of the axe head over (the logic being you're creating the socket for the axe handle, which irl are never circular on axes).  Bend is a special finishign move that both moves and finishes.  You apply it to an orange square, it removes that square, and creates a finished bend above ('up' from) it.  Bend is also the only finishing move that can itself be moved.  All other finishing moves are permanent, and if the player mis-applies them, They either have to SPLIT them off, or melt the piece back down and start over.  The following graphic shows that the bend moves can be applied to previous bend spots, moving them upward to square, and creating the new bend below them.  Bend is the most complicated move.  DISH and SWAGE simply apply themselves to the tile they are used on, just like DRAW.  In any case, with all BENDs and DRAWs applied, the axe is now finished.  It can either automatically pop off, or there could be an output slot.

Following are some recipes for other basic tools.  I'm also showing some of the moves that would be used for optimal effect, in grey (the finishes do cover up some moves that would be required beforehand).  The optimal original ingot position is shown in red outline. 

You can see that the hoe, pick, and knife use punch to create a hole for handle attachment.  Sword and pick use SWAGE to create rounded areas (quillions and pommel on the sword, the points on the pick).   Dish is used to make the curved area of the shovel.  It could cover even more area, and be used on hoe as well, if desired.  There's some balance to be kept, as some of the recipe suse a lot of moves (sword, shovel, axe) while hoe uses very few (could be upped with DRAW and DISH) and pick also uses few, despite being probably one of the most often-needed tools. 

The basic summary of the system is using a set of well-defined moves and rules for applying them.  Perhaps low tier anvils have a menu to ghost in the patterns for various things, to help players out in the beginning, while high tier metals demand the player remember what's going on (or use the wiki of course).  Or maybe no patterns to start even.  Maybe patterns are just orange and yellow square, without giving the symbols of the finishing moves, so at least the player is asked to logically remember the finishing moves associated with different areas.

Some complication could be added in the form of pritchel hole and hardies.  Punch command can only be used on an orange square directly over the pritchel hole.  The hardy hole would accept a bottom swage which would be required to use the SWAGE finishing move, or a wedge hardy, required to use the SPLIT basic move.  There could be an additional slot to the side to place the hardy in, and it could be an expendable item if desired.

GUI-LESS? - Now, to move the system GUI-less, the piece might need to be actually visible on top of the anvil, which would be incredible I think.  So other players could see each addition of a pixel to the pattern, as the piece progresses.  They wouldn't see finish moves show up though, just the basic pattern.  So you could walk by and tell that a player is smithing an axe!   Furthermore, the player might actually have to craft separate hammers for the moves, and use them on the specific pixels, kind of like cleaning leather in tfc.  Example hammers might be:  A lump hammer for HIT.  A 1kg sledge for UPSET.  A Cross-peen sledge for SPLIT (possibly with hardy).  Rivet hammer for PUNCH (or if pritchel hole is a thing, then punch might be 1kg sledge over pritchel hole.  Chase hammer for draw, straight peen for BEND (or add a horn to the anvil, and a 1kg sledge over horn creates BEND), ball-peen sledge for DISH, and swage hammer for SWAGE (possibly in combo with hardy).  And to top it all off, tongs.  These would be used in specific areas to rotate or move the piece.   Tongs could also be required to be in the hot bar to move hot ingots around, and to do any smithing at all.   Or, to avoid complexity, maybe they just have a 'blacksmith tool kit' that allows them to change hammer modes.   But the interesting thing would be, if each move have a tool, plus tongs, that's 9 major tools.  So a rl part of the player's skill, will be knowing what tool they want to make, and what hammers they need to make it.  They can get all 9 basic tools and pretty much fill their hotbar, making things more awkward, or they can know that they need a pick, and the only tools they need to pick up are 4 hammer, plus tongs.  

SKILL - Speaking of skill, a coded player skill could play into this system - and actually the existing system too - by affecting the likliehood of a move being applied.  So a novice, each time they try to apply a move to a copper ingot, maybe they have a 75% chance or something of it actually occurring.  But if the same novice tries to work steel, each move has only a 10% chance of actually happening each time.  So they'll use up a TON more hammer.  The final result could have durability penalties for each hit (successful or not) to discourage low-skilled players from trying to smith beyond their skill level.   Further, some of the less-used moves (punch, swage) could have lower chance to happen than other moves, to compensate for their not being used  much.   Successful hits could result in sparks flying.

So that's the idea, I hope I explained it clearly.  I know it sounds super-complicated when laid out in text like this, but I think it could actually be fairly intuitive, and I think people would appreciate that it simulates blacksmithing even closer than TFC (I'd argue).  If it could be made GUI-less, so people could see the piece develop, see sparks fly, I think they'd go wild over it. I haven't even dealt here with how the system might handle multiple ingots, sheets, multi-stage armor (get your DISH hammer warmed up!), etc.  I wanted to see if there was any interest in what I've described so far, before I got into more detail.  I know there's a desire to keep 'core systems' not too prohibitive, but I think this mechanic could be a showcase that people would wow over, maybe.  Would love to hear opinions. 

 

 

 

 

[Tyron]

Very intriguing system, thank you for the detailed explanation! I personally also always wanted to have some kind more visual, less button clicking system with sparks and stuff ;-)

Let's see how the other team thinks about it.

[Theunknownspecimen]

This system sounds complex to use, not really hard, just complex. Though I do like it or at least I think something like it would be cool. I don't have any expounding thoughts right now, but I will post if I get any.

[redram]

It could kind of be as complex or not as is decided appropriate.  When giving ideas I usually go for maximum complexity, as it's relatively simple to scale it back.  For instance you could just do away with the finishing moves, have no horn or hardy or pritchel hole.  Just anvil surface.  Then it's fairly similar to the clay and flint stuff from TFC, just with more squares, and a few rules.  You could discard punch, and get by with hit, upset, split, and the arrow buttons.  If you don't require things to happen over the anvil face, you could get rid of the up and down arrows.  You could maybe let the player switch the direction of the upset (repeatedly clicking on it?), and do away with rotation arrows too.  (Though come to think of it, in the GUI-less form you could do away with rotation by allowing upset to create the block away from the player, so they just change which side of the anvil they're standing on)

And, it could be a mixture using progression.  Perhaps copper and bronze tools don't have finishing moves.  Just the basic orange stuff.  Lets newbs get a handle for the system.   But beyond that, you add the finishing moves to the equation.  That might allow the player to get used to the first set of moves before adding the next layer of complexity.

[Saraty]

Sounds very complicated for me but im very willing to learn this. I think beginners "pattern" highlights should be definitely available at all times for optional use (player could freely enable and disable it). I know for sure, that my dad will need it.

Question: In VC hot ingots are cooling down slowly. I assume that would also be included in this system right?

[redram]

Yes, I was thinking the overall pattern would have color gradations that would change as it cools down.  So you might have to remove it and reheat it, and it'd then have to store its status.  It'd be pretty incredible if the current state of the ingot could be reflected in the inventory icon, but it might just have to be a 'worked ingot' icon which the player has to place on the anvil to see where it actually has been worked to.

[Saraty]

Simply 'worked ingot' sounds better to me.

Ok, You have my "yes" for this suggestion if the beginners help will be included. 

[Ivan]

I love the idea Redram, and it this sort of forging mechanics is something I have thought about before. But seeing it laid out in such detail (well done btw) I now have reservations. It seems to be mainly be an exercise in time, rather than skill. What is created in the end will not differ no matter who is doing the forging. A sword will still be a sword, a hoe a hoe etc etc. Or am I missing something?

To create the feel of real metal work might be a simpler exercise; and manipulation of a hot ingot (lots of sparks etc) onto an anvil, then depending on which part of the anvil the ingot is placed, an item is produced. So it is mainly an exercise in producing the right visuals and sounds to get an authentic feel. 

Also it should also be noted that working of copper and bronze is rather different than the process for iron etc, Bronze items were cast rather than forged. So there may be room for variation in metallurgy, without the complications (or as players might call its: frustrations).

Maybe the system you have laid out in such detail could be reserved for players who want to try master craftsmanship, where the effort/time would pay off.  

[redram]

As far as copper and bronze casting vs smithing, that's fine in terms of rl, but I think, but it's a game of abstractions, and I think it's fine to be able to smith those metals.  If you have both options, it gives you the option to have an easy cast blade (of lower quality perhaps), or you can smith it for extra time and raise your skill and make it better quality, so then when you get into actual smithing metals, you're not a total newb.  I think this is good game mechanics.  To me any time you make the player choose between two paths, you make the game more interesting and engaging, vs having just 1 path that requires no strategic choice on the player's part.

As far as skill, at the veeery end of the suggestion I had a bit about skill, but I'll expand upon that here.  To me there's two kinds of skill: rl skill, and game skill.

The idea I mentioned earlier was for game skill, that being that each pattern will have an optimal number of moves - the minimum possible.  If the player gets it in this many moves, they get the best quality.  Quality might affect durability, damage, weapon speed if they have such.  But, the player's code skill might have an effect on whether each strike actually takes effect.  So, even though the player might know exactly how to execute the fewest strikes, each strike might only have a 75% chance of causing the needed effect.  So, on the average they will produce a 75% quality weapon, despite knowing the exact best pattern.  As their game skill goes up, the chance of each strike taking effect goes up.  As i mentioned, this could be a sliding scale based on metal tier.  So an Expert smith might have 100% bronze strikes and be able to always produce the best, but their steel strikes might be 75%.  So they can choose to create perfect bronze stuff, or less perfect steel stuff.  The chances could be broken down even further, so base metal moves have 75% chance at X skill, but finishing moves at the same skill on the same material have only a 65% chance.   Even individual moves could have different percentages for the same skill level and same material.  All depends on how complicated everyone wants it to be.  

Here's my thoughts on complication: on the one hand, complication might create barriers to entry for certain players.  If they feel overwhelmed watching a let's-play, they may not buy the game.  It's a valid concern.  On the other hand, some people (like me) love extra detail.  The more detail to the process, the more scope you have for tech improvement.  For instance, having the code-skill affect the chance for a move to take, you now have the scope to create machines that increase that chance.  So you can create a low-tier knock-mill trip hammer.   This trip hammer has a much higher chance of executing moves than a novice smith.  So early game, the player builds this to help them out.  It becomes useless later when the player's skill is on par or better.  But then you have better trip hammers for higher tier metals.  But perhaps trip hammers only work on basic moves (orange in my example).  For finishing moves, the player must build a planishing hammer (this might be only for upper tiers, when the player has combustion or steam power).  You could drill down and have tools that only affect specific moves, like a roller mill for making sheets, that works only on the HIT move (presumably sheets would take tons of hits).  Down the line you have die stamping machines, where you put in an ingot, and in one move out comes a perfectly orange shaped piece, needing only the finishing moves.  Put a shaped orange piece into a roll-former, and it automatically recognizes the shape, and applies in one move all the bends necessary.  The drill press automatically recognizes and applies all punch moves.  Such machines would apply more than one to the total moves of the piece though, so that hand-smithing remains still the best method in terms of quality.   Now you've got the player building all these stand-alone tools, and having an actual metal shop.  Imagine if the player had enough tools to justify a line-shaft setup!  Players would gobble that stuff up, I think.  In the current mechanic, where is the room for such advancement?  The simpler the mechanic, the less room you have for advancements in tech, imho, because there are fewer variables to improve.  But, to help out newbs, is why I proposed that copper and bronze might not require finishing moves at all.  In fact, even higher tier stuff could have the option to just stop at the orange moves, and get a lower quality tool.  In the end, the whole game is a time sink of one degree or another.  But if something takes a lot of time, then tools can be invented for it to take less time.  The more helpful machines and such, the more goals the player has to accomplish.  And accomplishing goals makes players feel good.

Going back to skill, lets talk about rl skill.  This is something that TFC does well with it's system.  The player has to not only have a good idea of how the moves work, to find the target, but in order to be most efficient, they must be good at juggling hot ingots without melting them.   This becomes important when cast-forging an anvil (welding all the double ingots as you cast them, so you don't even have to make or use a forge, or spend fuel re-heating the ingots) or draining a crucible and making the most efficient use of the residual heat from the casting, so as not to waste more time and fuel than absolutely necessary on the forge.   It's a very elegant system, very complicated in execution, though not in principle.  But the fact of the matter is that irl, some people just can't handle it.  Some can't get past the finding of the target.  Others can handle that part, but do not use their time and casting heat efficiently. 

This system removes half of the rl skill element in the form of random numbers and targets.  The heating balance is kind of still there, but sort of not because the ingots don't melt, so you can probably just blast everything full heat and not worry.  Instead, this system will encourage rl skill in remembering the patterns.  They are much larger patterns than tfc flint knapping, and some people had trouble just with the knapping patterns.  So I think there will be those who are good at remembering patterns, and those who have to use the in-anvil supplement, or use the wiki.  Naturally those who remember will be faster.  So the rl skill required will be, I think, much less than TFC.  But this is sort of a strategic sacrifice to, I assume, attract a broader player base, by removing the frustrating (but rl skill-requring) elements of TFC forging.  In the end, it does become an excercise in executing a pattern.  But so is the currently planned system.  A much simpler pattern, and the mechanic is frankly not engaging at all.  It will however be the easiest for the widest array of people.  To me there's a point at which one might want to stop dumbing things down, as I think there's a large crowd that craves more complexity.

[Jdbener]

I... what is this, fourth? fifth? the motion for this feature. This sounds like the type of thing that would set Vintage Story aside from the competition.

[morsmage]

I like this idea as well. Complex is fine as long as it doesn't draw out the process for no discernible reason and doesn't put off new players (or make them have to go back and forth to a wiki).

As a less complex version of this, you could mix this idea with an idea (lifted partially from a mod called MineFantasy for MC). You heat the ingot, place it on the anvil, then use a GUI like this (also like TFC knapping) to "draw" the shape of the item you want to make. Then you simply exit the GUI, and "hit" the anvil with a hammer several times. Sounds and maybe visuals would let you know how close you are to finishing or when your ingot has cooled too much.

The thing I like about MF/TFC/VC/VS anvil mechanics is that it takes time but isn't simply a progress bar. It feels immersive and believable within the world.

Also, I agree that GUI-less crafting is nice. I am also a fan of requiring players to have several things placed nearby to work with. A true crafting area would have many tools placed. In this case, a forge, a bucket of water, and an anvil.

[Tyron]

With drawing the shape morsmage probably means a large crafting grid when you can place the items in a certain arrangment. I see now several ways to go.

In increasing complexity:

1. MF-Style: Placing items on a anvil crafting grid and simple hammering it
2. VC-Style: Choose a recipe then follow a number of techniques
3. TFC style with a simplification: Using a number of techniques to line up 2 markers, but instead of requiring to hit the exact mark, allow be in a ertain  range of it. The closer you get to the mark, the higher quality the tool would be.
4. Redram style: Crafting with or without a crafting grid

Any other promising systems out there? Perhaps we should make a poll for it.

[Jdbener]

2 hours ago, Tyron said:

Perhaps we should make a poll for it

I have a subtle fealing that you will get a lot of votes for #4 

[Saraty]

Poll is a good idea!

[morsmage]

@Tyron 

Quote

With drawing the shape morsmage probably means a large crafting grid when you can place the items in a certain arrangment. 

To clarify, I was suggesting something more like a TFC knapping/Redram style grid. I don't think that you should need multiple items (ingots) to create a single item. 

[redram]

Glad some folks liked this idea.  I haven't really had time to consider or address multiple ingots in detail.  My examples all worked ok for single-ingot items, but once things start to require multiple ingots, things get more complicated.  I can think of a few routes. 

SUCCESSIVE SMITHING - One would be a TFC-like route where ingots are successively welded together, and made into sheets.  Creating a slowly additive process.  In this scenario, the starting shape would presumably vary depending on what is put in.  So an ingot is a 4x8 block of squares, double ingot 8x8.  But then there's the question of sheets.  If the player has to expand double ingots into maybe a 12x12, to make a sheet, then it would be logical that placing a sheet down to work on would start with the same 12x12.  That leaves rather little working area around the edge though.  Also, if the rule is that all moves must occur over the anvil, either the working area will need to be significantly larger than 16x16, or the pieces will need to be able to hang outside the working area.  If you start with sheets for armor, then presumably armor would involve a lot of bending, punching, splitting and dishing.   Little to no swaging or drawing, and maybe not as much hitting and upsetting, since the sheet would probably already be close to the end size (for breastplates) or maybe even larger than, for helmets or boots.  So that's one route, successively welding and making double ingots and sheets and maybe even double sheets, and smithing them into the finished armor.  The successive combining would dictate the multiple ingot cost.

STACK INPUT - Another option would be to have an 'input' slot or a few.  Here the player puts the number of ingots needed, and maybe any add-ons or other things that might be needed (leather, rivets, buckles, flux).  In this case no matter how many ingots the player has in the input slot, the starting shape is just one 4x8 ingot, because it'd just be problematic to start with multiple ingots in the smithing area I think.  This will obviate the need for successively combining into sheets and double ingots, if that is not desired.  It would also require many more hits and upsets than would starting with a much larger sheet.  Which would make for more work, for the larger piece.   It creates a couple problems though.  One, is unfinished ingots.  If the player has put in 8 ingots for a breastplate, works them awhile, and then has to take them out to re-heat, how is the progress stored?  You probably would not make 8 individual unfinished ingots, as that would get messy if the player did not put them all back in the anvil (due to confusion, or whatever).  I'd assume the very first strike that the player takes, the 8 ingots would be combined into one single unfinished 'ingot' (If flux is to be a thing, this would also instantly use up 8 flux from the flux add-on slot).  So unfinished ingots would need to not only store the moves progress, but the number of ingots that are combined into that one unfinished ingot.  And this would need to show for the player.  Because if they didn't make a lot of progress before removing to reheat, it may not be totally obvious what they were on their way to making.  So if this is the system, maybe the ingot icon in the work space needs to show a number, indicating how many normal ingots it contains.  This means that if the player gets confused, and makes 6 ingots into a shirt shape, they're out of luck, as that's not enough - so they'll need to be able to melt the 6-ingot 'worked ingot' back down into 600 units of metal.  But conversely, if they use 8 ingots to make a pants shape, are they allowed to get the pants?  Even if they spent too many ingots?  Could the player be allowed to use more ingots than required, and get a durability bonus?  That would be interesting.

MULTI-PIECE - A third option, would be for armor to be a combination of smithing and crafting grid.  This is probably my favorite system.  In this system, the player does not even make a helmet, pants, or a shirt in the smithing area.  They only make plates and chain mail.  They then arrange those in the crafting grid to get the final product.  This acknowledges that any armor involving plates of steel, also involves flexible chain areas, and leather belts to hold things in place.  So, for a breastplate, the player might make four armor plates, three chain maile, and two leather belts.  two armor plates would be the should guards, and would be in the top right and top left.  Two more armor plates would be the front and back breastplates, and would be in the center and bottom middle.  The three chain would be left, top, and right center, and the leather belts bottom left and bottom right.  Assuming each plate and maile costs 1 ingot, the final cost would be 7 ingots and 2 leather.  Pants could be top middle plate, and two plates on each side center and bottom, maile top left and right, and leather center and bottom center.  Cost; 7 ingots 2 leather.  Or remove bottom right and left plates, if cost absolutely must be less than body.     This would also open up a route for chain maile armor (all maile and leather).  You could bump the chest recipe a bit by having the front and back breastplates actually be specific pieces, made from a double-ingot.  That would bump the chest cost up to 9 ingots, but the player could still start with a 8x8 double-ingot, so that we wouldn't have to deal with how to start with smithing a plate.  Or you could just require 4 plates and 5 maile for the plate armor chest, no leather involved.  So then just straight 9-11 ingots.   Or, there could be a special armor crating workbench (so the player can't just do this in their in-person crafting grid).  That workbench could have auxiliary slots for accessory items like leather straps and buckles.  So then you're free to make all 9 grid slots be metal, and yet still involve some leather.  You'd also tend to localize armor crafting to a workshop, and have another piece of furniture for the workshop.  Both good things imho.  The one problematic thing about this method is, it doesn't have an obvious spot to incorporate flux into the routine.  Which may or may not be a bad thing depending how what plans there are (or are not) to have flux in game.  You could make all armor plates start with a double-ingot, so then the flux would be taken at the point of forming the double ingot.  You might need to reduce the armor plate cost of some of the above recipes (which would allow for more maile or leather).  Or, armor plates could use up flux at the point the player takes them from the output slot.  It's not the logical place for it to happen, but it would allow armor plates to be single-ingot, and still have a flux cost.

It this multi-piece scenario, chain maile may require special consideration.  Making wire is very different from other smithing processes.  It opens up the possibility of having the player make a series of drawing dies and having the pound hot ingots through these dies in succession.  This would of course later be greatly sped up by machinery, perhaps a mid-tier machine that is just one die, but still the player pounding the ingot, and another which is a machine where you put in the hot ingot and it outputs wire (having such machines would also conveniently lay the groundwork for copper electric wire, in the advanced form of the game).  So the player gets wire from this process. There there could be a mandrel required (also used for barrel hoops?) to shape the rings.  or a smithing recipe, that starts with a bunch of long single-pixel strips, that must be split and bent.        Or if all that is too much work, just have the player start with an ingot, expand it to a large size, and punch out a ton of alternating holes to make a checkerboard pattern, and that produces chain maile.

So, I hope this has given some more food for thought on how this system might work in a multi-ingot scenario.  I'd hate for everyone to get on this bandwagon based on my simple initial examples, and then find out the multi-ingot stuff isn't to their liking.   I'd like to make some graphics to illustrate some of what I'm talking about, but I just don't have the time right now.  If anything was unclear, just ask, and I'll try to be more clear.

[Tyron]

Saraty and me talked long and hard about Smithing mechanics today. I think we've managed to finalize a concept worthy of implementation.

Redrams first chapters describes a great first implementation, so we'll skip any of the finishing moves for now to begin with that alone as proof of concept that is easy to understand for players and faster to implement. We'll also add one more rule to allow the use of multiple ingots/plate.

First of all there'd be 2 types of ingredients. Ingots and Plates. The latter can be crafted from the first using the anvil as well. Any recipe either can only be made from plates or only made from ingots. 

When using an ingot, the player will start out with a 3x7 area of pixels and with a 9x9 area when using a plate.

For each additional ingot/plate added to the current crafting session the player gains an additional contingent of "pixels" (e.g. 30 for ingots, 10 for plates but cannot mix&match), which are used up for each pixel player expanded using "Hit" (0-8 pixels cost) or "Upset" (0-1 pixels cost)

Each applied technique that is beyond the shortest path + tolerance will cost your certain amount of durability loss on the final tool.

This should satisfy most of the requirements 

[Erik]

On 8.4.2017 at 5:11 PM, Tyron said:

Each applied technique that is beyond the shortest path + tolerance will cost your certain amount of durability loss on the final tool.

Is this really a good idea? The system you describe would be easy to learn and easy to master, which in my opinion wouldn't be perfect. There only would be one way to perfectly smith something and you could theoretically look it up at the wiki.

So here's my idea: Each smiting product has three smithing stats: sharpness, hardness and mass (names for stats are subject to change). Each one of these smithing stats has a value between 1 and 100. These smithing stats interact (using simple formulas) to create the different tool stats. Durability would result of hardness and mass, cutting and piercing damage from sharpness and mass, blunt damage from mass, swinging speed from mass, mining speed from sharpness and hardness, defense from hardness. Different tools would require different tool stats, which require different smithing stats. The formulas for tools stats are also tool specific, the mining speed of an axe is more influenced by sharpness than by hardness, a pickaxe's mining speed would be influenced more by hardness than by sharpness.

So how do you get the smithing stats? Mass accounts for the amount of metal in the product and starts at 100 and gets less when the player executes a split. Therefore the player can lower the mass by splitting until every additional "pixel" is used. Hardness is can be archived by a hardening process applied to the ingot, so hardness can only be archived before the actual smithing. I'm not yet sure, how that hardening process will look, but I think that only steel should be able to be hardened. Sharpness can be archived by sharpening on a sharpening stone or with a sharpening kit, it is only possible after smithing. The itself sharpness is determent by the time of sharpening, the sharpening device used and maybe other factors, it shouldn't be possible to reach 100 sharpness by using just a sharpening rock. The sharpening uses a bit of mass and hardness to increase the sharpness. When using the tool, the durability AND sharpness gets lower. You can regain sharpness by sharpening all the time, but it would cost mass and harness, therefore also raising the durability loss rate and lowering other tool stats. 

Too complex? Well, the exact smithing stats are hidden, so they don't confuse players. What the player can see on the tool:

Weight (mass):  Light|Average|Heavy   (1-30|31-69|70-100)

Hardness:         Soft|Average|Hard       (1-30|31-69|70-100)

Sharpness:       Dull|Average|Sharp      (1-30|31-69|70-100)

A bit of guesswork is required, skilled players will have some advantages. The tool stats are depicted with exact values. And now some possible formulas, note that different materials have different material base stats:

Durability: MaterialDurability * [0.5 + (0.9 * (mass%) + 0.1 * (hardness%))]; Durability is initially set after the smithing, when it reaches zero the tool is broken and can only be repaired by being turned back into pure metal (Metal returned: Metal used for creation * mass%). It also translates into the loss of sharpness when using the tool.

Durability loss rate: MaterialLossrate * [1.5 - hardness%]; The durability loss rate describes the durability loss per use of the tool. It has a value between 1 and 10.

Swinging speed: ToolBaseSwingingingSpeed * MaterialSwingSpeed * [0.5 + mass%]; The lower the swinging speed, the better. It has a value between 0.5 seconds and 5 seconds. ToolBaseSwingingSpeed describes the base swinging speed of a certain tool type (sword, axe, etc.) and is has a value between 0.1 and 1.

Mining speed: MaterialMiningSpeed * [0.5 + (ToolMiningSpeedDestribution*sharpness%+(1-ToolMiningSpeedDestribution)*hardness%)]; Mining speed describes how long it takes to break a block. ToolMiningSpeedDestribution has a value between 0 and 1 and is tool specific.

Blunt damage: ...

Cutting Damage: ...

Piercing damage: ...

I think you get the idea. Every tool stat except Durability is recalculated, when smithing stats changes (using the tool and sharpening).

Would like to hear your feedback on this System, admittedly it's more about tools than smithing. 

 

[Tyron]

If I understand it it right though, your proposal is just an extension of the system I described right? That would be the point of implementing something simple first, and then add onto that some more challenging or less deterministic mechanics.

Generally I think there are more suited mechanics to make the smithing more interesting that does not involve that much number tracking. Or perhaps I'm just not seeing the whole picture of your idea, you've written it a bit vaguely. I can't quite picture a step by step process - and as you say this is more about tools than the smithing system itself. There's gonna be quite some more things to smith beyond tools.

[Saraty]

On 4/9/2017 at 9:12 PM, Erik said:

Different tools would require different tool stats, which require different smithing stats.

I like when tools require extra work and specific smithing or finalizing actions to improve their stats. After completing basic smithing milestone, improving the process by adding more complexity to it is always something i'm looking forward to.

[Tyron]

Now as I'm in the middle of an example implementation I stumbled on one big design question: How can players now what shape they have to smith? The smithing grid is very large compared to the 3x3 crafting, so much harder to memorize, and I really don't want to turn this into a game of memory.

The only alternative option that comes into my mind is add a recipe selector that in turns highlights the outline you need to craft, and I'm not sold on that idea at all. Any input is welcome.

[Jdbener]

15 hours ago, Tyron said:

Now as I'm in the middle of an example implementation I stumbled on one big design question: How can players now what shape they have to smith? The smithing grid is very large compared to the 3x3 crafting, so much harder to memorize, and I really don't want to turn this into a game of memory.

The only alternative option that comes into my mind is add a recipe selector that in turns highlights the outline you need to craft, and I'm not sold on that idea at all. Any input is welcome.

 

Patterns that can be crafted in the 3x3 grid? Including a pattern material (paper maybe?) which when crafted together in the 3x3 properly will get you an item which you can place on the anvil to highlight the outline you want.

[redram]

On 4/30/2017 at 11:15 AM, Tyron said:

The only alternative option that comes into my mind is add a recipe selector that in turns highlights the outline you need to craft, and I'm not sold on that idea at all. Any input is welcome.

That's basically what I'd assumed would happen - an outline, or ghosting of the relevant tiles.  Is there any particular reason you're not sold on it?  If it seems too easy, what about a pattern selector which perhaps doesn't put the pattern directly on the anvil, but puts a small picture of it somewhere, for the player to refer to? Big enough they can discern the 'pixel' jaggedness at the edges, but not gridded, so not super-easy to count?  That should provide guidance, while removing some of the 'paint-by-numbers-ness' of an on-anvil outline.  Or, there could be on-anvil tile ghosting, but only the tiles immediately adjacent to the existing metal?  Then the player won't be able to see the entire pattern at once, and they might do some inefficient moves, only realizing as more pattern is exposed their mistake.  Over time they would probably learn the most efficient moves though.

There was also the notion that low tier anvils could provide more explicit guidance (on-anvil-outline maybe).  Mid-tiers could provide a small picture or adjacent ghosting, and high tier could have no help - you're a master by then right? (there's always the wiki as backup)

[Tyron]

On 5/1/2017 at 9:24 AM, Jdbener said:

Patterns that can be crafted in the 3x3 grid? Including a pattern material (paper maybe?) which when crafted together in the 3x3 properly will get you an item which you can place on the anvil to highlight the outline you want.

Interesting but i'm afraid that still requires you to select the right recipe in some way and clutters your inventory, 

11 hours ago, redram said:

That's basically what I'd assumed would happen - an outline, or ghosting of the relevant tiles.  Is there any particular reason you're not sold on it?  If it seems too easy, what about a pattern selector which perhaps doesn't put the pattern directly on the anvil, but puts a small picture of it somewhere, for the player to refer to? Big enough they can discern the 'pixel' jaggedness at the edges, but not gridded, so not super-easy to count?  That should provide guidance, while removing some of the 'paint-by-numbers-ness' of an on-anvil outline.  Or, there could be on-anvil tile ghosting, but only the tiles immediately adjacent to the existing metal?  Then the player won't be able to see the entire pattern at once, and they might do some inefficient moves, only realizing as more pattern is exposed their mistake.  Over time they would probably learn the most efficient moves though.

There was also the notion that low tier anvils could provide more explicit guidance (on-anvil-outline maybe).  Mid-tiers could provide a small picture or adjacent ghosting, and high tier could have no help - you're a master by then right? (there's always the wiki as backup)

Yea ok, since I also don't have no better idea i'll do just some basic recipe selection for now. It's merely supposed to be a proof-of-concept anyway. Thanks for the feedback u 2

[redram]

I feel like the smithing system has had a pretty good run in it's current form.  Not a lot of complaints.  I'm SUPER glad it was able to be made GUI-less in world, and it's really awesome that two people can work on the piece at the same time.  I had thought that this was kind of what the devs decided to take from my original suggestions, and leave it simple, but Saraty suggested she was still interested in some of the other stuff.  So viewing smithing as still a system in progress, I thought I'd revisit some of it in light of the actual mechanics we have now, and touch on some things talked about in discord.

FINISHING MOVES

Finishing moves.  With the in-world system we have, it's not really practical probably to use symbols as I'd originally envisioned.  The 2x2 voxels are a bit small, at least for the types of symbols I had.  So, there's probably a couple other options for finishing graphics - namely voxel thickness and color.   Basically, allowing the player to ad half-high voxels (if that's practical code-wise) and changing the color.   Color could be problematic in that there's a lot of metals, each with their own colors that gradate through the heat stages.  So you'd have to pick colors that stand out on any metal. 

The original finishing moves I'd suggested were draw, bend, dish, and swage.  swage was probably the least used of them, so maybe we discard that.   Draw was basically for edges, so that's probably one to keep.  Dish and bend are similar, yet sort of different.   

In consideration of all this, I think you could basically distill it all down to raising and lowering pixels, if that mechanic is practical, and not mess around with color.   The player lowers the blade edges, and also any dished middle areas.  The player adds additional voxels on top for raised parts. I think most tools provide opportunities for each.  I think that lowering or raising more than a single pixel (1/2 voxel) in height may not be practical.  It can already be hard to distinguish voxels in the glowing metal, and too fine of gradations would exacerbate that I think.  Plus being smaller than one pixel creates visual texture inconsistency I think.  I think it would be neat if any voxel that is being raised or lowered could be darkened a bit compared to it's neighbors.  That would help distinguish it, and working metal does often cool it (darkening the color) irl.  Once this has happened, the player can no longer use that voxel to move metal around.  Maybe you can add or subtract a half step to get it back to 'normal' again and keep moving it.  Or maybe you just have to melt it and start over.

You could also have different colors of voxels for different moves, or really simple symbols, like a slanted line for draw, a line for dish, a chevron for swage.  But those will appear very artificial.  I think ultimately the above paragraph may be the best option.

HAMMER MODES

Copygirl expressed a dislike for the hammer modes in discord.  I don't know if it's maybe worth revisiting my original suggestions about move-specific tools?  So then you just use the tool, and don't need to select a mode.  There is of course a danger of it becoming too complex.  But consider this:

You could have the basic hammer just move 1 voxel at a time.  You *get rid of* the heavy hit move, which creates voxels surrounding itself.  I've found that in practice, I only use this move a few times on the sword.  Now granted, I don't smith axes (200 units vs 100 cast?  not thanks) or shovels, because again casting is just quicker.  So I don't know how they compare.   But basically I think you move the heavy hit function to the trip hammer, down the tech line.  So it would be more a tool for really large items perhaps, like sheets and armor plate.  Normal tools might just be more practical to do on the anvil, using the 1 voxel at a time hit.   Point being, that gets rid of one tool mode, and hence one tool in a tool-only scenario.  Leaving your basic anvil with just hit (1 kg sledge) and cut (chisel in combo with sledge).  This leaves room for another tool for raising .5 voxel (straight peen hammer), and another for lowering .5 voxels (ball peen hammer).  This would match Milo's Discord suggestion of trying to limit the tool number to 4.   Personally I don't subscribe to that, I think the player with want to keep multiple tools for backup, and will have more than one rack anyway.   Plus the chisel I would assume would double as a stone-working tool, maybe even wood working as well if that becomes a thing.  So it's kind of a wild-card tool maybe.

As such, I'd also just resurrect the possibility of tongs.  These could be used for removing hot ingots/tools from the molds and dipping them in water to cool.  And, if we don't want the player to have to constantly move around the anvil, they could also be used to rotate the pattern, allow the sledge to simply always move metal in the same direction relative to the character (toward or away probably).  player right clicks on the right side of the pattern (relative to themselves) and it rotates it clockwise.  Left side, CCW.  Shift-right, picks up the pattern.   Picking up hot ingots and such without tongs burns the player and forces the item to be dropped.   I can definitely understand if this whole tongs thing is a bit too far for some people, but I personally think it'd be ok. 

BELLOWS

There's also some discussion of bellows. I was assuming that was being omitted as too grindy, but if they're going to be a thing, I'd like to suggest they wear out over time.  One of the issues I had with TFC was that leather had very little use (and VS faces the same issue right now).   Your bellows lasted forever and aside from leather armor, quivers, and saddles, there was nothing else to do with it, so leather was eventually a useless material.  Leather does dry out, crack, and fail, and being next to a hot forge does not help it.  The more uses for leather, the more the player must make interesting choices in what they use their leather for.

TOOLS SIZES

Currently only the axe (and shovel?) use two ingots.  This creates a severe imbalance with casting.  I think that forge ingot requirements need to match the casting ones. 

WELDING

I also thought welding was a conscious omission.  Apparently Saraty likes it though, so here's a suggestion for a possible mechanic for that.  You have to take a piece of flux, and apply it to the pattern you've smithed so far.  Each time you right click on an in-progress tool voxel with the flux, it covers that voxel in white on top.  This is the flux.  You have to cover the entire surface of your in-progress tool.  Then you right click the next ingot, and it gets added to your voxel queue.  Now each flux drop fluxes 9 voxels - the target and the 8 immediately around it. This keeps the flux required to a sane number, and makes it so it could be beneficial to smith in a way such that you use the least flux (not smithing out super-long pieces, but trying to go in square-ish groups of 9).  This adds a little interesting variable.   It would mean that it takes more than 1 flux to weld an ingot, but depending on how flux is made, how hard it is to get, maybe that's not such an issue.  I forget how many voxels you get per ingot, but at 9 per flux, I think it would be 4 at most?   I certainly never felt like I had flux problems in TFC, once I found it.