Technology System Version 5.3

[Richard Bruns]

The previous technology thread was getting close to 150 posts, so I started this one to continue the discussion. I wanted to put a nice summary of the current tech system plans here, but we are in the middle of discussion and the post count went up much faster than I thought it would. So I'll just have to wait until we get things hashed out.

By the way, do I get a prize for starting our 300th thread?

[Beör]

quote: Originally posted by Richard Bruns on 09-19-2000 07:58 PM The amount of tech transferred in the migration depends on the number of people...

This is great. I really like this.

quote: Modeling at the EG level would give exactly the same tech diffusion result as modeling at the civ level.

Disagree. If tech level is modelled at the EG level, it would of course not be the same as the civ tech level.
In 2.2.1 (civ tech level in any area is equal to the maximum of the tech level in this area for all constituent EGs) it would almost invariably be considerably lower and never higher than the civ tech level.
In 2.2.2 (civ tech level in any area is equal to some other function of the tech level in this area for all consituent civs (probably geographical distance, 'cultural distance' and social status (discrimination) would enter into these equations)) tech level of one EG might be higher than the civ level, but would then probably be lower for most other tech areas.
So the maximum tech level that the receiving civ should be able to receive when based on the EG tech level will be lower than when based on the civ level.

quote: The ideal tech level of the conqueror is compared to the actual tech level of the conquered square.

So you are in a way modeling tech level based on squares. Why not go all the way and model it at EG-level - it will be much smoother I promise .

Seriously: I don't quite understand this ideal tech/actual tech part. Please elaborate, and be very specific about at what level they are modelled, and how they are derived - and while you are at it think about how migration and conquest fits in!

quote: By the way, thanks ...

You're welcome. I think it is rather fun. Just hope nobody will be pissed off, because I'm such a nitwit .

BTW Please take a look at the piece I posted in the Model Organisation thread. I think it is very relevant to this discussion as well.

[F_Smith]

Richard:

(In reply to you post in the last thread)

If your version of 'tech level' is the best thing your civ can do/build, then it's a collection of knowlege, by definition, correct?

And if 'tech level' is a collection of knowledge, then it must be collected/stored somewhere, correct?

So then what type of object collects and stores a civ's knowledge information? An 'infrastructure' object, like a University, business, etc, correct?

For example -- the best bridge a civ can build is determined by the engineer businesses and govt engineering agencies in that civ. Businesses and govt agencies.

Infrastructure objects.

At least, that's how it seems to me.

[Richard Bruns]

This is getting out of hand.

First, I learned that axi's new infrastructure model will not use a formula that I had been relying on. Without the actual tech level A, the tech model is incapable of applying itself to the econ model in any good way. This means that the model is severely crippled.

Without A, all of my explanations about ideal and actual tech level are shot to pieces. So to answer Beör's question, I don't know anymore. The change in the econ model has destroyed my vision of the interaction between tech and everything else.

Then, several core assumptions used in the tech model are being systematically demolished. The model is built on the idea that "knowledge" is something that is an integral part of the civilization as a whole. It is an abstract thing. It was not meant to be stored in a building in some province, and it was not meant to belong exclusively to certain people. It was meant to be a quality that belonged to the entire civilization, not some specific part of it.

Basically, F_Smith and Beör are trying to stretch the model to cover things it was never meant to cover. They are trying to turn knowledge into a physical, concrete thing when it was originally an abstraction. By insisting on extreme realism and erasing the foundations of the model, they have come very close to destroying the model. It was not designed to be flexible, so it is breaking as people attempt to stretch it.

And then F_Smith seems to claim that the model structure is a horrible monster that will wreck the game and drag innocent programmers to their doom, spreading chaos and devastation in its wake as the army watches helplessly. . . His solution is to somehow create "components" out of the big system. I fear that this will be impossible, given the fact that I designed the model as a single unified procedural system.

I'm almost ready to abandon this model and build something new from scratch.

[axi]

Richard: NOOO!!! STOP! IT'S ALL A MISUNDERSTANDING! There is nothing wrong with the tech model! Give me some time to compose a comprehensible answer and I'll explain!
-----------------------------------
It is my fault actually. Look what I posted in the infra-thread:

quote: Richard: No, feed-in from the tech model to the infra-model does not involve A. A will be useful for applying the effects of infrastructure and it will probably use the #of_units_input/#of_old_units to do that, but you won't find that in my spreadsheet. What you'll see is a "tech mod" vector, a "decay" vector and two commodity cost tables that need to be figured out somehow from the tech model. That what was my second paragraph referring to. I can't make it more simple than that.

Now look what I meant to say:

When I am talking about feed-in from the tech model to the infra model, I am not referring to A, because a will be used in the "Infrastructure Effects" area, which is not modelled in my proposal. So A doesn't fit anywhere in the infra spreadsheet, but it fits very well to the game we want to make! # of units and A will be used to calculate the effects of infrastructure, like it happens with the production function, where A is the tech level of K (Kapital units).

Of course this approach means that each infra unit gets better as tech advances. This is what partially causes the need for a tech mod. If we want, we can take A out of the effects calculation and put it, in exactly the same manner, in the commodities cost calculation. This way, the value of each infra unit does not change, but it's cost is constantly reduced. In other words, it is a matter of definition of what a "unit" actually is.

As for the claim that the tech mod is actually a social thing, it is irrelevant; it has to engulf only technological determinators. Are there any; I think there are. Today, although are communications are far more efficient than ever, we spend a much greater amount of our resources for them; they have become indispensable to us and technology is the sole factor behind this effect.

I am rather sleepy now; I will continue clarifications tomorrow.


[This message has been edited by axi (edited September 20, 2000).]

[Beör]

Wait a minute - don't panic here.

I hope that Axi's post has cleared some misunderstandings regarding A (whatever that is).

I must say that I am certainly not trying to make the tech model do anything it was not supposed to. I tried to analyse what happens if we model tech level at different levels of the game hierarchy. And two things in particular will cause problems: migration and conquest. These need to be adressed - and this was my way of provoking a discussion on the subject. I am sorry if I stirred things up and made your life generally miserable, Richard .

I must go to bed now.

I hope tomorrow we will be able to resume the rather rewarding discussion we have had over the last week.

[Richard Bruns]

axi: Okay, sorry I misunderstood you. I still am not comfortable with the prices, decay, etc. It seems to violate the "black box" principle. I'll see if I can come up with anything.

Beör: Don't blame yourself for saying that the emperor has no clothes. Sometimes it takes a "nitwit" to see things like that. Your analysis was accurate. When held to certain standards of realism, the tech model fails. We either have to accept those failures of get a new system.

By the way, I'm not upset, annoyed, or miserable. The system has been proven inadequate, and I can live with that. You all have proven that I didn't design a realistic or accurate system, and I know that that the system cannot be scaled to provide that accuracy. I still think that it would work well at a certain level of abstraction, but if the game environment gets detailed it will probably fail. That is a symptom of a design flaw. We could keep the game modeling confined so as to cover up that flaw, or we could try to come up with something better.

[Richard Bruns]

F_Smith: Where is this tech model you proposed earlier? I did a search but was unable to find anything about a tech model other than Harliquin's and LGJ's. Until recently, I had never heard of the existence of any other tech systems. I'd like to see what you had.

[F_Smith]

Richard:

If you go back and re-read the threads "Technology System Ver 5.x", it's all there. The real meat begins in the second 'life' of that thread, Technology System Version 5.1. About the middle of the page I offer my version of 'technology' system analysis.

quote: Axi's tree is very close to the system I'm more interested in, with a few caveats -- [*]I still feel that the level one 'techs' are actually 'theories', and should be handled at the 'civ' level as a 'theory object', not as a 'technology object' (an entirely different concept, since a 'theory' is an imperfect understanding of a 'tech' which should approach true understanding of that 'tech' as the theory improves), and not at the tech object level (a tech object will likely have methods for turning this into that -- iron and tin into steel, that sort of thing.) I think this design choice will affect us much, later. For one thing, a 'theory' object should be allowed to improve 'skill' objects in more than one of your 'fields of knowledge'. A better military 'theory' can and often does, in fact, lead to changes in society. Changes in societal theories can lead to improvements in Tech 'skill'. That sort of thing. [*]The level 2 'techs' are, to me, not really 'groups' in any real way, unless handled as civ 'skills' -- another completely different object with an entirely different implementation. Art (fine jewelers, say) doesn't directly improve law; Math doesn't directly improve Biology, brilliant tacticians don't directly improve weapons, and on, and on. If this is a civ's 'skill', that would be one thing, but then it must be handled and coded differently than as if it's a tech.[/list] Just to reiterate, I'm willing to go as is. I'm not the strongest analyst you'll find (I sometimes go back to code weeks or months later and look at what I've written and wonder what I was on that day . . .) But -- I think ya'll need to refine the analysis and design of this model, or face many of these same issues later, during coding. Designing this as a single tree lumps theory, skill and 'application' objects(the actual 'technology' part) into one object, mislabeling things that aren't 'techs' as 'techs', unless I'm missing something entirely. In the RTS I'm also working on, this is the model I'm using. I think it's the only way to go. Just my pinched pair o' pennies.

* * *

I would refine and change a few of those objects now, in hindsight, but that was the basic OO 'analysis' process for a 'technology' system.

Two objects -- a 'theory' object and a 'knowledge' object.

[*]Tech Theory is potential knowledge. Theories would be stored at the 'civ' level in 'infrastructure' objects.
[*]Tech Knowledge is stored in ethnic groups, and is used to perform tasks. Knowledge objects typically take inputs of 'raw materials' and 'tools', and produce some sort of output ('steel', '+1 to happiness', etc).
[*]Each knowledge object would also store it's own 'skill' level, so that when a group of people 'use' a knowledge, their (and only their) 'skill' using that (and only that) knowledge grows.[/list]

Both can be passed to other civs/ethnic groups, according to their own 'turnhandler' methods.

At least, that's a first pass. I welcome any discussion or debate on the analysis. I almost always make at least one big mistake when I first do something. It's in there somewhere, I'm sure . . .

[Beör]

These discussions regarding the nature of things, the proper way to model the game in OOA terms, and the level at which this should be done are proceeding in so many threads that it is almost impossible to keep track off.
So since this is really a reply to F_Smiths post I will post it in this thread. If anyone feels that this or any of my other posts belong in a different thread feel free to copy them there (but please tell everbody about it).

While a theoretical analysis of the real world may reveal the 'true' objects we should concern ourselves with, I think (and I know Mark has made this point earlier) that it is important that we also think OOA/D from the game perspective. There is no need to reveal complex object structures, if the effect on gameplay is negligible.

Here are some question I asked myself:

Who am I?
I am 'controlling' an empire. This is maybe the greatest abstraction of all. I am a Being in charge of the destiny of an empire. I have everlasting life (or at least live for the longest possible duration of the scenario), and in some areas considerably more control over what is going on than my mortal reallife counterparts had historically - although I have no 'magical' properties. This longevity and limited omnipotence permits longtime planning and is actually the prime instrument in achieving better results than historical counterparts, whose aims rarely exceeded their lifetime. This is not very lifelike, but it's what makes the game fun.

What do I have to work with?
I have two basic kinds of ressources, that translate into all that is happening in the game: Land (handled as mapsquares), which is immobile and in practice unlimited, and people (handled as EGs), who are mobile and unlimited. Actually there's a third ressource: Time (handled as game turns), which is limited. It is a basic assumption that people given the time have an inherent capability to work the land and produce a rudimentary output. The output takes the form of units (food, production, service, specials).

What is the empire?
Another huge abstraction: In a way I am the empire. You might also say that the game-interface is the empire. I control the interface, thus I control the empire. The empire/interface is linked to one of the basic ressources of the game: land, (since it's immobile). By controlling land, the empire controls the people living on the land. So if people leave the territory under my control, I loose them. This translates into: If the game-interface does not have control over some land inhabited by some people, then the game is over. Only under very special cirumstances (nomadic civs) can an empire exist without land, and this obviously requires special handling.

What do I aim to do in the game?
Basically I want the empire to survive - the game to continue until scenario end (this is more than most historical civs have achieved). On the other hand surviving as two hunters and a gatherer in New Guinea might be easily achived, but not a very rewarding game experience. So there has to be more: Something like having a large empire with a large population. There might be other qualifiers like having a happy population, clean environment, total world dominion, immense wealth etc, but for now let's concentrate on the simple aim: I want a huge empire with lots of people.

How do I improve my conditions from the starting ones?
This is were all the fun is.
The trick is converting the output from the people working the land into more people, more land, or increased productivity. This is basically done by creating infrastructure. Units are transformed into infrastructure, which in turn either makes it possible to get more people in the already controlled territory (increasing reproduction, decreasing mortality, increasing immigration, decreasing emmigration), extends the controlled territory (military units), or makes the people more productive (anything from a plough to a nuclear power plant and beyond). Infrastructure extends the capabilities of your empire beyond the basic rudimentary level.

Is that it?
Basically this could be it. Notice that not once have I mentioned tech/knowledge/research/lore or whatever terms are used. Tech is not a goal in itself, only a means to an end. And notice - contrary to what F_Smith stated - that it is infrastructure that perform tasks. Infrastructure increases land, people or productivity by converting units into some sort of bonus - knowledge doesn't.

Where does tech fit in then?
Tech makes it possible to differentiate infrastructure in time and space. It is a qualifier for infrastructure. It makes some infrastructure more effective than other. The people working the land has an inherent capability of transforming some of the produced output into RPs which are then converted into tech. The tech in itself does nothing, but it permits the construction of more sofisticated infrastructure, which in turn increases land, population, or productivity. You may have the knowledge, but if you don't have the necessary infrastructure it won't do you any good.

So infrastructure produces, knowledge does not. Are there other distinctions between the two?

Infrastructure is immobile - it is a physical thing existing in a mapsquare - it belongs to the land (mapsquare). If people leave the land, the infrastructure will still be there. They cannot take it with them. I have a small problem here with simple applications as the plough, which could probably be carried. The infrastructure model talks of mobile parts and immobile parts, but what is really done here is the disassembly of infrastructure into its origin Units, which can then be converted into infrastructure at a new location.

Tech/knowledge is mobile - it is not a physical thing - it belongs to the people (EG). If people leave the land, they take their knowledge with them. Knowledge is not left on the empty land floating in thin air. Knowledge cannot be separated from the people. You can sell a copy of your knowledge, but you cannot sell the knowledge itself.

Infrastructure is volatile. Infrastructure requires upkeep to counter wear and tear over the years. You have to pay a certain amount of maintenance cost.

Knowledge is persistent. You do not have to maintain it. When it is - it is. I know that here I am at odds with the present tech model. When - in real life - it seems that knowledge deteriorates, IMO it is because there is no need for the associated infrastructure. As long as infrastructure associated with the tech exists, the knowledge will persist. If the last piece of related infrastructure disappears, the tech is obsolete and therefore it seems as if it has deteriorated. But I claim that if need be the tech will still be there. You might have to dig a little for it but it would be there. I do not think that it is necessary to discover the same thing twice.

Infrastructure can only be produced at a cost. You have to spend Units to get infrastructure - it does not grow by itself. It does not dissipate or reproduce.

Knowledge can and will frequently be achieved for free. FE by doing things, having contact with other EGs or using applications knowledge increases automatically. Knowledge can reproduce itself indefinitely, dissipating to other EGs. It is possible to spend Units on increasing knowledge, by investing in knowledge producing infrastructure, but you cannot 'purchase' new knowledge from nature. You can, however, purchase already discovered knowledge from another civ.

So I guess that what I am proposing is that a mapsquare object holds one or more EG objects and one or more infrastructure objects. An EG object holds one or more knowledge objects of varying type and level.

Infrastructure objects function like F_Smith suggested in 2: They take some input and produce some output. If the infrastructure is destroyed, but the EG remains in the square, it will always be possible to build the infrastructure again.

Knowledge objects permits the creation of infrastructure objects: The presence of certain levels of knowledge objects of a certain type in any EG in the square will permit the creation and maintenance of certain infrastructure objects. If the EG having the permitting tech leaves the square, the infrastructure will gradually be destroyed, due to lack of maintenance.

This is the clean, mapsquare view of knowledge. There is another possibility. Square-based knowledge objects could be aggregated into knowledge objects at a higher map-level, FE province or civ. How they should be aggregated is at the moment beyond me.

Similarly, infrastructure objects could be aggregated at higher levels providing province- or civ-wide improvements. Again the details elude me.

EGs will of course also be aggregated, I refer to my post in the Model Organisation thread .

I am not sure that this will help y'all - but it sort of cleared my mind.

[F_Smith]

Boer:

I almost totally agree.

[*]One thing I disagree on is the infrastructure doing the producing. I'd argue that it always (at least, until AI is developed) requires people to produce something. And they 'use' the 'infrastructure' to produce, as a 'tool'.

'Factory Workers' would use a 'factory' (along with raw materials) to produce output. The factory doesn't do that itself.
[*]The other thing I disagree on is your not seperating 'theory' from 'application'. That's why I'd have 'theory' objects that are stored in certain 'infrastructure' objects (churches, universities, govt agencies, guilds, etc). The 'theory' objects would determine in part what 'application' knowledge an ethnic group could learn. If they have access to a church that 'teaches' literacy, they could learn to read and write.

There's also a matter of 'skill level' at using any of these techs. I think it's one of the most fun parts of this game to have an ethnic group who are the most skilled boat builders in the known world, and that kind of thing.[/list]

Other than that, I think you've outlined the basic program data architecture for gameworld info.

[Richard Bruns]

quote: F_Smith and Henrique: What exactly do you want? . . . If you could provide specific examples and explain exatly what you are interested in, we will comminicate better. Now, I have no idea what you want to do.

quote: Ya'lls system should play very much like the one in 'Masters of Orion II', which is fine. It'll work. I don't think it's the best, but it will be fine. --- I just so much more enjoy the StarCraft/Age of Empires/Civilization/Alpha Centauri approach behind the scenes, personally. And the industry seems to have migrated to that, away from pure numbers. They are much more fun, in my experience

F_Smith, we were not rejecting OO design. We were rejecting your insistence on the old way of doing the tech tree. Anyone who has made a civ2 mod knows that prerequisite based systems are a huge mess. They are coomplicated, micromanagement intensive, and extremely hard to customize. And it seemed like you were simply being a die-hard supporter of that type of system.

That way of modeling could never even hope to tackle the issues that are starting to injure the current tech system. The civ2 tech system cannot deal with migration, conquest, or EG's in any reasonable way.

The current tech problems were not foreseen by anyone. The thing that the current system has trouble with is migration, conquest, and society. F_Smith, you never said anything about those issues. In fact, you continuously said that the tech system would work fine. You simply claimed that the civ2 system would be more efficient and easier to produce, and you were right. Rehashing an old, tired system is always easier than venturing out into the unknown.

It seemed like your primary focus was to defend that old system. Nobody else saw a person who was trying to help analyze and code the current system or work on developing a new system. We saw someone who was rejecting innovation and telling us that we should use the old civ2 modeling.

Beör: Your ideas and analysis are coming too fast for me to deal with. I need more time to think about what has been said. Please bear with me; I'm trying to keep up.

[Beör]

F_Smith

You are so right on your first point:

People produce

Infrastructure catalyzes and amplifies production and diversifies the output. Instead of just producing food, production, service, specials and RPs, infrastructure makes it possible to produce a wide range of possible outcomes: From military might to happiness, from health service to oppression etc.

Knowledge qualifies infrastructure making some infrastructure more effective than other.

I am not sure I understand the distinction between theory, application and skill. I see them all as knowledge objects - or at least derived from a common knowledge object. I am actually really fond of the level 1, 2 and 3 techs in the present system. What are the differences between your system and the existing system.

One thing I know I don't like: A knowledge object belonging to an infrastructure object. IMO infrastructure cannot possess knowledge in any form. Knowledge is possessed by people. If knowledge belongs to infrastructure it would be tied to the land with several consequences I don't like:

If the square is conquered the knowledge passes to the conqueror and is lost to the original owner. While it is true that knowledge might be conquered, I will argue that the conqueror only takes hold of a copy of the knowledge (the copy might even be imperfect, incomplete or misleading). The original knowledge will remain with the original owner. Knowledge persists.

If the infrastructure object is destroyed the knowledge would vanish into thin air. There is a historical corollary in the burning of the library of Alexandria. While many priceless scrolls disappeared, I doubt if this had anything but a transient effect on society due to loss of knowledge. It was a disaster, but because of the artistic value, not the knowledge content of the burnt scrolls. Art is infrastructure and can thus perish in a fire, knowledge persists because it is with the people.

If the people choose to migrate they would not be able to take the knowledge with them. Need I repeat myself ?

Your example of a church teaching literacy is IMHO nonsense . Just as a factory cannot produce anything, a church cannot teach anything. People produce - in this case the clergy produce knowledge - catalyzed by the church. The clergy possess 'literacy' and the clergy are part of the people.

I know/guess that what you want is a kind of civ level maximal tech level that limits the levels of more practically oriented derived techs. In a way this is similar to the 'Ideal' and 'Actual' tech levels mentioned by Richard I guess (I have trued to find info on this without success). But making a separate object is not the way to go. As with our other base objects this should be achieved by aggregating the tech objects of the EGs in the civ. Don't ask me how yet - I haven't figured it out.

Am I right or am I right? (sorry that should have been 'wrong' - just slipped )

[F_Smith]

Richard:

I also mentioned I preferred 'tech trees' to this abstract approach.

But then, as in the post Boer and I are discussing, that I quoted above, I offered an OO analysis that was very close to what you had, just broke the different levels of tech down into different objects -- 'technology' subclasses, if you will.

I think at the time my use of OO terminology made me impossible to understand.

So now that you understand OO terms, what do you think of my OO analysis of your tech approach? Can that work with your system? Can you design game rules to work with that OO storage hierarchy that you would like?

* * *

Boer:

I would agree, both might extend an abstract 'technology' object. But that is actually a plus.

I basically took their system and did an OO analysis on it, which is why it looks so similar. By breaking the monolithic tree into the two 'theory' and 'application' objects, and combining them with 'infrastructure' objects, I am able to reproduce the tech system they want, and more.

As far as 'infrastructure' objects 'owning/storing' tech 'theory', here I'm talking about written or stored knowledge. Books in a church library, for instance. Knowledge in a teacher's/priest's head. Knowledge in a plant manager's head, operating procedures for a factory, Etc.

As in the Library at Alexandria, yes? Certainly you'd agree that the library should 'store' or 'hold' book knowlege (theory objects), yes?

This provides the same levels as the monolithic tech tree, with none of the restrictions.

[Beör]

I really love this.

With the wink of an eye we've (well you've ) OOed the tech model and maybe even the infrastructure?

Although I still strongly disagree with having infrastructure objects possess knowledge objects.

Yes, libraries contain many books, but they are only digested, organised reference to something that is inside a persons head. The books are infrastructure, not knowledge. If the books disappear, the knowledge remains either as other copies of the same books or inside the head of people, on microfilm, on CD or in songs and tales.

The knowledge inside a preachers head - well there's knowledge, but a preacher is not infrastructure, he's a person. He's part of the religious class, remember. The preacher is not part of the church object, neither is a teacher part of a university object. If the church/university is destroyed, the preacher/teacher will still exist. Same goes for the planter. Just as knowledge objects cannot belong to an infrastructure object, neither can people: They are separate entities with independent existence.

Operating procedures in a factory is part of the infrastructure. They do not become knowledge unless they are in such general terms, that they apply to other similar facilities, in which case they would not belong to a single piece of infrastructure, but be a part of general human knowledge.

The fact that knowledge could be destroyed or conquered along with the infrastructure or the square is very counterintuitive to me.

The loss of the library of Alexandria was a catastrophy - because of the loss of art (infrastructure), not because of the loss of knowledge.

Let me turn it around: In gameterms, what do you aim to gain by having knowledge objects belong to infrastructure? What is the purpose?