Hey everyone thanks so much for trying out the mod! If you have any suggestions or comments regarding gameplay or balance please don't hesitate to comment here or to make a thread about it. I want to hear from you all what your experience has been so that I can know better what direction to take things or what might need to be tweaked.

Pretty cool!

What triggers the recipes to appear? Perhaps when they enter my inventory? In multiplayer will it be when one player has in their inventory?

Suggestion: maybe some mid/late game recipes to get more or less of certain kinds of ore/oil. Or perhaps that was intentional to use promote more interesting fish generation based on what ore has too much?

What is the strategy for which recipe to make fish with? Are all the fish-generating recipes net positive? I’m assuming more advanced items convert to even higher rates of return on fish?

Early game is a bit limited by landfill, but still enjoying the challenge! :)

What triggers the recipes to appear?

Yeah right now it's just whenever a player has an item enter their inventory, the unassemble recipe for that item is unlocked for the whole force (team).

What is the strategy for which recipe to make fish with? Are all the fish-generating recipes net positive? I’m assuming more advanced items convert to even higher rates of return on fish?

Higher level items do have a bonus to their fish return based on 2 factors: the number of unique ingredients needed to make the item, and the "complexity" of those ingredients (basically how many steps there are in making it).

The recipes that have had one or more of their ingredients turned into fish are in general going to be the worst for fish returns because for each ingredient that is replaced, the "unassemble" form gains the replaced ingredient as a result instead of fish.

The resource generation is something I've given some thought to because it can be a little difficult to get resource automation off the ground as it is. I like the simplicity of the current system and the balancing aspect is neat, but I do get the sense that this is a point where things will change a bit in the future.

After a quick (in factorio standards) look

Does pollution do anything? If not, you may turn it off as default in the mod. Someone will be doing a fishy megabase:p (edit: looks it is already changed)

Do the resources spawn on islands intentionally?

Bigger landfill stack. The landfill production is decent, but the whole inventory of landfill creates not that big of an island. Seablock increased it to 1000 I think.

In Helmod already unlocked "fishy unassemble" recipes are not visible without enabling "show locked recipes". But it shows all fishy recipes (and all unresearched ones).
No unassemble recipes show in sandbox mode with unlocked tech. Not even after creating items.

Doesn't the green science feel too easy? It requires only fish. Others feel similar. It needs tons of fish, but the whole logistic chain is much shorter.

I have found a nice recipes iron ore-> fishes. Both fish-efficient, and space-efficient. I have also found reasonable recipe to turn stone into fishes, with less efficiency but I can see it as removing byproducts, and iron->fish (good to balance the loop) . But I can't find efficient way to turn/sink copper. The best from early-ish seems to be the decider combinator, that eat 2 times more copper, and is one of better (IMHO) fish producer.

I didn't reach a moment when it would be critical, but the main challenge in the later game seems to be balancing resources, or getting rid of the excess. (edit: int he current path all resources has their own recipes, so no longer a problem... nor a challenge)

Fish production is one thing, but I also were looking for recipes that bypass normal production. For example, by creating and unassembling yellow underground belts we could create yellow belts (3 belts per 5 iron plates).
But it looks like I overlooked two rebalances, so my recipes rather stopped working, time to restart:)

Edit: I just restarted the map, and got power armour mk2, worth 16000fishes. Looks like a too large bonus before the player unlocked automation:)

Thanks for taking the time to write this out. It's really funny that you got a mk2. Each item has a 1/31000 chance to get looted from fish, which I just kinda added as an easter egg. I'll probably remove it later or make it more rare.

As for anything recipe related, it's important to know that the whole thing is procedurally generated from the seed provided in settings. I've mostly tested on the default seed so far (because that's what most people are probably gonna play) so I can relate to your points on specific recipes. I'm working right now on a some pretty significant tweaks to how the value of each item is calculated, which will probably help at least a bit to make recipes that had all their ingredients turned to fish a bit more difficult to make (fish costs will be higher on average).

One thing to note about the individual fish -> resource recipes is that they're significantly less efficient in terms of space and efficiency than the one that gives 2 of each resource, so hopefully the challenge of balancing will remain a large part of the mod (and if not I can always do some more tuning).

Speaking of tuning I'm still in a phase of very actively playing around with changes to try and get the balance right, so I apologize for whatever degree that gets in the way of base planning. I'm just as new the the mod as anyone else, and getting my head around how exactly everything comes together balance-wise has been surprisingly difficult.

Fish production is one thing, but I also were looking for recipes that bypass normal production. For example, by creating and unassembling yellow underground belts we could create yellow belts (3 belts per 5 iron plates).

The "alternate recipe paths" thing has been very much on my mind but I haven't been sure how to balance it yet. I'm still working on getting fish production right and since that's tied up very closely with the goal of viable alternate recipe paths, it's been on the back-burner.

Thanks again for your interest in the mod I really do appreciate it, and I'll continue to work hard at making it feel good to play.

Hmm, 0.7 dropped, so it is time for the late feedback from 0.6.2 ;-)
That version felt much more grindy. The early loops were also much less efficient. But I did not go too far this time (like two of those https://ibb.co/PhChZ7Y :))

A warning: the seed used by the mod changed in 0.7, but when I was starting the new game, in the mod setting it still was 12345, only pressing the small red button reseted it to 1234 (the new default I think) (Edit: oh, changing the seed restarted the game... to rewrite recipes. It make sense now. But if it is possible to force the change of the default seed, it is probably a good idea )

BTW, the loops are very sensitive to to changes.
Lets say a production line that produces X fish and has a "cost" (number of factories it the best I think, or power... but power is just fish) C.
But is also consume Y fish (either directly or through resources). Then, the cost per fish is C/(X-Y)
Y is often in the early loops 0.5 - 0.6, so, in a sense the changes in recipes are 2, 3 times bigger than they aperar at first glance... I'm not sure where I'm going with it. :)

Yeah, the sensitivity is part of why it's been so hard to nail down the balance, but I think the new algorithm in 0.7 is pretty solid; most recipes are very close to or below 1 plate =1 fish returned. From the limited testing I've done with 0.7 so far it seems like the early game is still pretty "grindy", but it should in theory pick up substantially when you get to advanced fish processing.

One big change I made (and actually forgot to put in the patch notes) is that you get 100% return on the non-fish items when unassembling instead of ~50%. I'm not totally sure yet, but in theory this should makes it viable to generate higher level materials though fish-negative unassemble loops (for example a loop where you're effectively paying fish to generate steel). I think this would be pretty cool but we'll have to see whether this feature breaks the game, or remains totally useless. Hopefully it'll land somewhere in the middle.

I found a nice, compact, very early game fish positive loop without ores ;-)
https://ibb.co/PGK3669
Make underground belt, disassemble it (fishes balance themselves, eats iron), disassemble belts (enough iron to close the process, free fish)

Is there a simple (explainable:)) rule, how many fish is in a recipe?

I'm sorry I might have just broken your loop with the update I just put out; there was some weirdness with items that are made more than one at a time.

Is there a simple (explainable:)) rule, how many fish is in a recipe?

The system iteratively evaluates the value of each item based on the value of the items required to make it, taking into account raw materials cost, craft time, and complexity of the recipe. After that is done, it goes though each recipe in the game and transforms 35% of the ingredients randomly. The fish cost is then the sum of the "values" of each item that was transformed, rounded.

Nah, still works, the underground belt loop did not changed, and the fish output from belts dissasembling dropped by 10%. The setup from the pic went from 50fish/min to 45.

I like the idea that not-fish ingredients cancel out when assembled and disassembled. It should allow mentioned item "transmutation" (like iron->belts).

@system. I think there is a bug. An underground belt is worth 5 iron plates + 2.5 belts (2 und = 10 Fe + 5 belts).
2.5 belts was replaced by 4 fish in assemble recipe, and the 5 iron plates was replaced by 4 fishes.
But shouldn't disassembling recipe give back 2.5 belts? Or the recipe should take 2 belts, like the assemble recipe (probably easier to implement).

The loop is possible and so effective because we get 5 belts back from one underground, while the original recipe used 5 belts for two
(as I understand, the intended resource multiplication should happened when fish is turned into raw resources, and the other recipe should be more or less "fish stable" ;-) Maybe with a bonus for higher tier items)

Looking for steel "transmuting" recipe I have found great recipe for steel. One fish can turn into... 10 steel plates! It is also smaller and less power hungry (if we count energy for smelting) than the original recipe (smelt all the way from ore). https://ibb.co/Y09WPdN
It looks like the reason is the same.
The original recipe for reinforced concrete is 20 concrete + 8rods + 1 steel + water -> 10 reinforced concrete.
Fishblock recipe exchange 1 steel for 5 fish.
But disassemble recipe is one reinforced concrete to 2 fish and one steel. We again, get 10 times more steel than expected!

It looks like the disassemble recipes do not take into account the number of items created by the original recipe (at least for normal items, 20 concrete +8rods looked more like 20 fish, than 2 fish, so fishes are divided properly.... perhaps:))

I was right about to go to bed when I saw this message and wow what a huge oversight on my part. I'll fixed it so that regular items can be fractional as well as the fish and now I'm going to bed. Nice find though! I love that you're getting to know the mod so well and thank you providing such great feedback.

It looks like the insomnia path worked (yep, I started writing this post a week ago;-) ). For now, it looks like one Fe/Cu plate is worth one fish (I'm, not sure how this appears from the initial assumption resources are 0.45 fish in the code :) I'm missing something), at least for a couple of items I have checked.

" Craft time for basic resource generating recipes increased to 3 seconds from 1 second."
Now to feed a smelter columns we need an almost as big an area for constructors as for furnaces.
48 stone furnaces (or 24 upgraded ones) need 30 assemblers.
But is another incentive to switch to the "fish processing" recipe. Probably a good idea.

Since we are talking about fish processing. IMHO the output is nice. It eats and produces 1/4 of blue belts and designing a compact system (for example for 4 blue assemblers) that takes all the resources without throttling is a nice exercise (the output fluctuates, so some king of buffer is needed).

Stone seems to produce one fish per 2-2.5 stone (not 1~1 like ores). Is this intentional? The 5/2=2.5 ratio seems to remind the ratio between ores and stone in previous version of fish processing.

Some thought on recipes (1234 seed)
Using basic oil procesing and making it into solid fuel generates 405 fish per 9000 oil. Sinking plastic (9000oil+202.5coal) gives exactly the same 405 fish. It may be partially beause different routes (do you now rate item from the cheeapest recipe?). Coal seem to be worth ~0.5fish (from disassembling granedes). So it is also less valuable than ores (just liek stone... are those related? Both were buffed in fish processing output).
In this seed coal is not an ingreedient in coal liquefaction:) Feeding it hish gives more or less the same petroelum output as from fish processing (both with cracking and advanced oil), but fish processing create also ores.

Again, science. Fishy recipes allow to skip part of the production. In 1234 for science it means bleu and yellow science (and of cource red, green and military) can be made without oil at all, only utility science need sulfur and plastic (that one can be replaced with fish).

TL;dr: the main question: isn't coal and stone valued too little in recipes? I can turn ore to fish ~1:1, but stone and coal is less than 0.5fish per resource. Or that was the plan?

The biggest factor in stone and coal being worth less is that they don't have the intermediate step of smelting. My algorithm has items gaining value for every second it takes to craft them, as well as the number of steps the item takes to craft from raw materials.

Right now I'm using the "lowest tech" recipe to calculate the value of each item, but I'll probably change that to cheapest later, because for example right now petroleum is worth way more than the other products of oil processing which doesn't make sense imo.

Again, science. Fishy recipes allow to skip part of the production. In 1234 for science it means bleu and yellow science (and of cource red, green and military) can be made without oil at all, only utility science need sulfur and plastic (that one can be replaced with fish).

I have an idea for a newer system of replacing ingredients with fish that will preserve more of the original complexity of recipes, this will once again change all of the recipes though so I'm holding off until I have time to test and know fore sure whether it will be more fun this new way.

The lowest tech recipe is probably less efficient and the later recipes will do it better. But I'm not sure is it bad outbo The alternative is taking the most efficient recipe, but this mean the resources would be worth less until the player unlock the best recipe. Useless fuel at the beginning or great fuel recipe later.

It looks like (I may look at it wrong, the last big loop in do_recipe_valuation_pass), in the case of multiple outputs, the value of the items is determined by dividing the cost of the recipe equally among all product types.
So, if a recipe produces 100 of one type of item and 1 other type item, the first one is considered 1000 times less valuable.
Advanced oil processing is a bit like that, it produces 55 petro, 45 light and 25 heavy oil. 25 heavy oil is worth (without complexity and time taken into account) 12.5 petro gas or 18.75 light oil. But here 25 HOil is worth the same as 45 LOil, so HOil is 1.8 times more valuable than light oil.
In reality values looks reasonable (heavy-oil 0.0687, lubricant 0.0857, light-oil 0.1022, petroleum-gas 0.1133), from printout what becomes "canon" when it looks like pertoleum takes value from basic oil, HOil from advanced oil, and light oil from... cracking. Interesting.

The another case is if the multi-product recipe creates the main product and some byproducts. Let's take a fictional recipe: 10coal -> 5 coke + 1 ash. 1 coal is 0.45, the recipe (with bonuses) is worth, lets say 5. Now, 2.5 goes to coke, 2.5 goes to ash. Coke is worth 0.5, ash is worth 2.5. But I do not see how to prevent this. Or, rather, I do not see an easy way;-)

The cost of the recipe is increased by 0.12time (sounds good, it will be divided by the number of produced items) and 0.05total complexity (sum of the ingredients' heights in the rercipes tree). More different complex ingredients increase the value, but it doesn't matter how many ingredients there is. A recipe that work in bulk would provide a smaller bonus.

If someone else wants a quick look at values, here is what 0.7.5 produces
https://docs.google.com/spreadsheets/d/1y5Aq8y26f1y-sDiMU341ZBHI72dMzh6vDegXfmVti_w/edit?usp=sharing