Vokarya's Workshop: Wonders

I have one last pair of Wonders that I want to post before I start pulling everything together and merging for a final build. This is a pair of road-building wonders: the Via Appia and the Golden Spike.

Here's my reasoning for including both of these:
  • I'm moving Mining Inc from Railroad to Explosives. I think Explosives is a better fit and gives Explosives a nice broad range of content attached. This leaves Railroad as a one-trick-pony tech (the Railroad improvement) and it needs something else; therefore, the Golden Spike.
  • I hate to do anything only once and prefer things in pairs, trios, or otherwise linked. I think Via Appia serves as a neat precursor to the Spike.

Via Appia's primary effect is to lay down a Paved Roads network connecting all your cities. This sounds like a powerful effect, but I think it's muted a little bit because of the relative sizes of empires in the mid-Classical Era and that it requires both City Planning and Construction to build. If you get Construction first and start laying down Paved Roads, then Via Appia will wind up not doing as much. Via Appia uses Python to count back and forth between your cities and lay down the shortest possible road path between them. After the Via is built, it provides a little culture and GPP as all Wonders do, and it also provides +50% trade route income, which I will admit is somewhat of a placeholder effect and we can find something else for it to do once we settle on what Trade Routes wind up becoming.
View attachment 378740View attachment 378741

The Golden Spike lays down a Transcontinental Railroad on completion. What this translates to is that it will identify the west-most and east-most points on the building city's continent and connect them by rail to the building city. The tricky part of this is that map coordinates aren't enough if the continent wraps around the map edge; I was able to code for if it wraps around one edge, but a very artificial continent (like an S-shape that goes off both edges) won't quite work; the railroad will stop at the map edge. On the other hand, it will create a globe-circling railway on a map that is almost all land (like the Great Plains script). I am very interested in seeing saved games on maps where the Spike doesn't quite do what is expected.

After the Spike is built, it carries the idea of "Manifest Destiny" and "taming the wilderness"; it increases Worker speed by 25% (I might increase it to 50%) and 50% faster production of Pioneers (since they are already available) and Workers.
View attachment 378742View attachment 378743
 
Re: Maginot Line: Units sound better to me too.

:goodjob: with the routes wonders!
 
Indipendence hall is enabling all government civics. But by the time you build it, you've already discovered all Government civics. I think we should change the effect to something else.
 
45°38'N-13°47'E;13383373 said:
Indipendence hall is enabling all government civics. But by the time you build it, you've already discovered all Government civics. I think we should change the effect to something else.

Funny, I just built that one. Just for Great Person scores. However, yes, agreed.
 
Well, I never really liked wonders with all civics enabled, so I quite like it the way it is: It does not break the game allowing civics that otherwise would be eras away (like the Swedagon Paya does).
BTW it also gives GPPs and -50% Anarchy, which is a good thing even on its own.
 
Well, I never really liked wonders with all civics enabled, so I quite like it the way it is: It does not break the game allowing civics that otherwise would be eras away (like the Swedagon Paya does).
BTW it also gives GPPs and -50% Anarchy, which is a good thing even on its own.

I remember one game when my continent had no access to religions for a long time. Therefore when that Wonder came up, I built it then got Atheist civic. My civ researched faster because of research bonus. So, IMO, this one Wonder only should be permitted to grant all civics. Because if a team has religion, it's actually more profitable to be religious until later in game so it's balanced.
 
45°38'N-13°47'E;13383373 said:
Indipendence hall is enabling all government civics. But by the time you build it, you've already discovered all Government civics. I think we should change the effect to something else.

I didn't get a chance to change Independence Hall after the removal of the Fascist civic. I'm thinking about changing it to allow all Rule civics. That would give you Technocracy for now and at least one other civic (Single Party, or whatever we wind up calling it) in the future.

By the time you get to the late Renaissance/early Industrial, which is where Independence Hall will be built, I don't think unlocking all civics of a category is overly-powerful. Shwedagon Paya is powerful only because it comes relatively early. Later on, you already have some choices, so adding more isn't so powerful.
 
At some point, we are going to be swapping the abilities of Plato's Academy and Great Library, giving the Academy the +2 free Scientists and Great Library +1 science from all Scientist specialists along with a +25% science bonus (which I think works best). This will be done once the <SpecialistTypeExtraCommerces> XML tag is added. However, a tag like this has a lot more potential than just one building. Here is a list of other Wonders that I am planning to modify with abilities granted by this tag.

J.S. Bach's Cathedral: +2:culture: from Priest. This is in addition to its current effects. Right now, Bach feels like a weak version of Notre Dame. This will make it more of a cross between Notre Dame and Sistine Chapel. On a second note, I also want to move Bach's expiration up to Mass Media. Wonders should get at least two full eras of effectiveness to compensate for the effort it takes to build them.

Brandenburg Gate: +2:culture: from Great Military Instructor. +1 XP for all units is nice but not always guaranteed to make an impact. This is a little something to help it out.

Bauhaus School: +2:culture: from Engineer, replacing the +1:commerce:. Bauhaus School is supposed to hybridize Artists and Engineers. The Artist works fine with +1 hammer, but I'd be much happier with giving Engineers culture instead of commerce.

Edison's Workshop: +2:science: from Engineer. This replaces the current +1:hammers: from Scientist. Edison is supposed to help Engineers, not Scientists.

Einstein's Laboratory: +2:science: from Scientist. The +2 Scientist that the Lab currently gives have a feeling of "it's been done", and I'd like to do something else.

World News Network: +2:gold: and +2:culture: from Celebrity. I don't find WNN to be all that impressive right now, as it gives free Press Agencies which have been buildable for a long time, reduces Anarchy length (when there is a good chance that Cristo Redentor already prevents it), increases Golden Age length (this at least is useful) and cuts down War Weariness by 10% (not very much). I think giving it a good bonus to Celebrities, which are not easy to come by, would also help.
 
I have the new tag working, and will include a working example with your suggested change on Bach's Cathedral. :)
 
Perhaps <SpecialistTypeExtraCommerces> could also be included to civics too? One example: Intolerant +1:espionage: for priests.

Such a tag already exists for Civics: SpecialistCommercePercentChanges

It allows fractional commerce changes for specific specialist types. I noticed it while adding the building tag.

Edit: Free Church uses it, for one example. +0.50 :culture: per Priest.
 
I'm giving the Wonders a look-over to see if there is anything else I want to do with them while I am instituting the new tags for bonus commerces from specialists. Olympic Games has the combination of No :mad: in City, +1 :), and +1 :) per 20% :culture:. But if the Games are already negating all unhappiness, is there any reason for the happiness bonuses? The only thing I can think of is that excess happiness cuts down on Revolution instability, but if we want Revolution to have more of an impact, then this becomes even easier to cut out. What do you think?
 
By my count, there are still 42 technologies that need some help in order to get above the 2-trick threshold needed for viability. 22 techs are one-trick, and another 20 clock in at one major trick and one minor trick each. Several of these 1.5-trick techs are fundamental building blocks that I do not want to get rid of (like Agriculture, Archery, Chemistry, and Computers). I've been cooking up some Wonders to help some of these techs.

First of all is Gutenberg Bible, to help Printing Press. Currently, Printing Press has the Printer as its major trick, and the +1 :commerce: from Villages and Towns as its minor trick.

Gutenberg Bible has one major effect: it spreads your State Religion to all connected cities of all players when completed. It doesn't matter which religion you have, but you do have to have a state religion to build the Bible in the first place. This is a Python effect that checks that the city is connected to the trade network and that your religion could spread there (either the other civ allows non-State Religions to spread or it is already running your State Religion). I tested it myself and it will not spread to cities you have not yet revealed.

View attachment 381315View attachment 381316
 
Maybe you already know this but any wonders with python effects (where the AI are unaware of the building effects) should have a strong AI weight and flavors set or the AI will likely never build it. By strong I mean 100-500 for AI weight and 10-50 for flavors.
 
Maybe you already know this but any wonders with python effects (where the AI are unaware of the building effects) should have a strong AI weight and flavors set or the AI will likely never build it. By strong I mean 100-500 for AI weight and 10-50 for flavors.

I did have a Religion Flavor of 10 on it, but I will up it to 15 and add an AI Weight of 100. I have never really worked with AI Weight before. Do you think all Wonders should have an AI Weight? My experience has been that a lot of Wonders don't get built right away by computer players and seem easier than they should for a human to scoop up, although it does vary from Wonder to Wonder; Pont du Gard and Piazza San Marco go really fast. The only Wonders with an AI Weight are Sun Tzu (300) and Ascension Gate/Nanite Defuser (1000 each).
 
I did have a Religion Flavor of 10 on it, but I will up it to 15 and add an AI Weight of 100. I have never really worked with AI Weight before. Do you think all Wonders should have an AI Weight? My experience has been that a lot of Wonders don't get built right away by computer players and seem easier than they should for a human to scoop up, although it does vary from Wonder to Wonder; Pont du Gard and Piazza San Marco go really fast. The only Wonders with an AI Weight are Sun Tzu (300) and Ascension Gate/Nanite Defuser (1000 each).

Basically when the AI evaluate the buildings value, they assign arbitrary numbers or values to most of the different attributes on buildings. This works because "what" the building does is defined in the XML, and so the SDK can see the XML values and calculate a "value" for the AI. These values tend to be between 50-200 for normal buildings (depending on how useful a building is in the AI's particular situation, it changes based on financial difficulty, empire and city modifiers, etc). Wonders tend to be a bit higher, in the 100-500 range due to the fact that their effects are usually fairly powerful.

So enter python-created wonders. The SDK has no clue about your custom effects on those particular wonders. The AI might see a wonder that has no effect and slap a "0" on as the value. (The value may be slightly above zero, if the wonder has XML effects beyond the python code). But it will be under-valued greatly. The AIWeight tag is the poor man's python fix for this, you can simply soft-code (vs hard-code, since it's in XML) an AI value for the building. AIWeight is directly added to the AI value of buildings. So if the AI thought the wonder was worth, zero, then added 250 AIWeight to it, the AI would think the wonder was worth 250 and might build it.

It's fine to use an AIWeight > 0 for buildings with python effects, its the only way to tell the AI that this building is useful (even if the AI is not quite certain how it is useful). Flavors are similar, but are only added to the AI value if the AI cares about its military, religion, etc.

BUT you do NOT need AIWeights on normal, XML defined wonders. The AI can evaluate these buildings properly. Setting AIWeights on normal XML defined buildings and wonders is bad. This is using a hammer to guide the AI behavior instead of fixing an underlying problem. But where python-wonders are concerned, AIWeights are the only tool you have, and when you only have a hammer... ;)
 
Basically when the AI evaluate the buildings value, they assign arbitrary numbers or values to most of the different attributes on buildings. This works because "what" the building does is defined in the XML, and so the SDK can see the XML values and calculate a "value" for the AI. These values tend to be between 50-200 for normal buildings (depending on how useful a building is in the AI's particular situation, it changes based on financial difficulty, empire and city modifiers, etc). Wonders tend to be a bit higher, in the 100-500 range due to the fact that their effects are usually fairly powerful.

So enter python-created wonders. The SDK has no clue about your custom effects on those particular wonders. The AI might see a wonder that has no effect and slap a "0" on as the value. (The value may be slightly above zero, if the wonder has XML effects beyond the python code). But it will be under-valued greatly. The AIWeight tag is the poor man's python fix for this, you can simply soft-code (vs hard-code, since it's in XML) an AI value for the building. AIWeight is directly added to the AI value of buildings. So if the AI thought the wonder was worth, zero, then added 250 AIWeight to it, the AI would think the wonder was worth 250 and might build it.

It's fine to use an AIWeight > 0 for buildings with python effects, its the only way to tell the AI that this building is useful (even if the AI is not quite certain how it is useful). Flavors are similar, but are only added to the AI value if the AI cares about its military, religion, etc.

BUT you do NOT need AIWeights on normal, XML defined wonders. The AI can evaluate these buildings properly. Setting AIWeights on normal XML defined buildings and wonders is bad. This is using a hammer to guide the AI behavior instead of fixing an underlying problem. But where python-wonders are concerned, AIWeights are the only tool you have, and when you only have a hammer... ;)

What about Projects? I've seen no AIWeight or Value but I've noted that AI very seldom build projects (mostly in modern and transhuman era, as International Monetary Fund, ITER Fusion Reactor, First Cloned Mammal, etc). I was wondering if there's something we can do to improve the situation; actually, I wonder where the problem might be because there might be some other background problem, i.e. poor AI performance from industrial era onward.
 
45°38'N-13°47'E;13412223 said:
What about Projects? I've seen no AIWeight or Value but I've noted that AI very seldom build projects (mostly in modern and transhuman era, as International Monetary Fund, ITER Fusion Reactor, First Cloned Mammal, etc). I was wondering if there's something we can do to improve the situation; actually, I wonder where the problem might be because there might be some other background problem, i.e. poor AI performance from industrial era onward.

That is unlikely to have any affect. The AI does not compare buildings to wonders or projects, they are selected independently.
 
Top Bottom