Technology System Version 5.1

[Mark_Everson]

F. Smith:

I strongly support Richard and LGJ on this... IMO we just simply cannot have a "connections" type model for Clash. For one thing, what you seem to be looking for is extremely contrary to Clash's the-player-is-the-leader perspective. What leaders really understand, if they understand anything about technology, are level three and level four technologies in the current model. The things that give their civilizations real practical power. Perhaps our nomenclature is a bit screwed up, but so what. When we show it to the players we will think of sexy names for all these things. The approach being potentially too dry is certainly a legitimate criticism. I think the only way we are going to practically address that is in playtesting. If it turns out to be too dry, I think it will be easy enough to put more excitement in this basic framework. If necessary we can make applications themselves helper techs for the level one through three techs. It would make the system less flexible, but would capture some of the issues you are concerned about.

Axi:

Your list certainly does have better coverage in some areas than the existing model. Thanks for putting it up. Hopefully adding in some of the stuff you part will make the model a bit better.

Richard:

Your patch on my effectiveness idea strikes me as extremely reasonable...

[Lord God Jinnai]

F_Smith:

But inventors do use theory, even at the most basic level of invention. Even if the 'theory' they use is wildly incorrect. And since a good research 'model' must allow 'inventor' objects to use 'theory' objects and 'known tech' objects (or technologies) to combine them into a new tech object, I feel this model falls short.
-----
What ur saying right here is that the government employees the inventors. This may be the case on occasion, but overall that's wrong.

Also ur wrong about them using theory at the most basic level. At the most basic level they use trail and error. You'd have a tough case saying that's using theory.


That is, in effect, my point. The 'Fields of Knowledge' -- the first 'node' in ya'll's 'Tech' system -- in fact belongs in a different part of the design. A civ's basic understanding of Math theory does not, to me, belong in the 'tech' design
-----
How so? Math is one of the most important factors in science, economy, some religions, everyday life. The increased knowledge of mnth allows us to understand more. Without a high level of math we'd not understand physics as well as we do.

I'm wondering if ur just mainly complaining about the "label" of tech for everything? If so, that's fine. Give us some specific remedies for naming problem, remembering that level 3 and 4 are simply names the comp will see, not the player.


. The entire concept of a 'Social' tech is, to me, unworkable and unrealistic. I see 'social organization' as a seperate, distinct object with it's own behavior. It can influence many things -- but it is *not* a 'tech', as I understand technology. The same goes for the other 'fields of knowledge', as well. They should not be simple percentages and modifiers, they should be objects with behaviors.
-----
There will be "objects and behaviors" as you call them, but humanities (as it is called now) and the fine arts do advance. There are more advanced philosophies than others, same for religions, etc. The differnce is these aren't "techs" as u like to say, they are "advances" and they are heavily culturally oridented. If ur culture has a set mind against freedom, things like human rights, democracy, etc. will be improbable/impossinle to get, even if ur 50% points above the min rq. While I'm at it there will need to be a way to keep from a "flood" of new humanities and fine arts advances in the wake of a cultural change, otherwise they player would get too much way too fast.

My understanding of 'player control' is that they will not have anything to do with the actual number-crunching, merely in setting 'goals'. And if that is correct, then ya'lls analysis, which sums up skill, theory and tech as percentages, is (for my game-playing tastes) far to dry. I prefer a seperate 'system' for each, with seperate 'objects' (each with appropriate behavior) -- a 'tech' system, a 'theory' system and a 'skill' system. They must inter-relate in the 'research', but be seperate systems
-----
I think Mark and Rich have said everything i can.

Also other than Rich's brief comment i wanted to know what people thought of the social advances part of the model i post and what could be done better.

[axi]

You know guys, what I expected as a reaction to my list was a barrage of cries, grumbles and complaints, asking one of the following questions:

• Why {tech1} is not included?
• Shouldn't {tech2} level up (down)?
• Shouldn't {tech3} go under {tech4}?
• Do we really need {tech5}?
• What is the difference between {tech6} and {tech7}?
• If {tech8} is included, why {tech9} is not?
• The level# techs, shouldn't they be more (less)?
• Shouldn't we split (merge) {tech10} and {tech11}?

This way we would improve the list and, having these fields of knowledge well defined, we could go on configuring the helper techs, the percentages the applications and finally the effects of each item in Clash. (It's a lot of work but don't worry, some of it is already done in the previous model.)

What I get instead is a discussion about the distinction between theory and technology (or something like that). In that matter, I have to side with Mark and LGJ. Consider the following:

There are many different levels of theory. An example from my field of expertees: "Analysis of thermal turboengines" is a theory, but also an application of "Fluid mechanics", which is a theory too, but also an application of another, more fundamental theory, "Calculus". In my faculty students are taught both 3 theories, so that, when working in industry, they will be able to handle the application "turboengines" in a scientific way, complimenting the work of the technicians. Very few fields of knowledge are purely theories or applications; most of them are both in some extent. If we also consider the various interconnections between theories and applications, the thing gets too complicated to describe by simply dividing the items into theory and technology.

What do we do? We make a tree of items that influence each other in an interactive way, without defining a theory-application relationship, but only implying it, by the nature of the items and by the structure of the tree. In reality applications affect the advance of theories as much as theories help advance applications and so it goes with the tech tree: The player should be able to advance with direct funding only items in level 4 and, gradually, also in level 3 (maybe even level 2 in some cases). The resources thus allocated not only advance the desired item but they also dissipate in order to advance higher levels, which can subsequently lead to other related level 4 advances as well. This represents not only the formation of theories to fit the applications pursued, but also the various interconnections between techs. On the other hand, knowlwedge is also created unconciously within a civ as a result of it's characteristics, it's preoccupations, it's mentality, it's relations with other civs, in other words from the civilisational environment. This is not under the direct control of the player although it eventually depends on his own policies and geographic conditions. These RPs flow from above in the Way Richard has described and place themselves chiefly on level 1 and 2 advances (sometimes even on level 3), leading in an automatic raise in all affected items. Due to the nature of the interconnections, these resources will flow downwards, resulting to applications.

Imagine the tech tree as a net pulled upwards from various knots simultaneously: the whole net is raised according to it's local connections with the knots being pulled. We could also parallel applications with fish caught in the net, coming to surface as the net is pulled. Theoretic research would then amount to pulling up the whole net, instead of trying to pick up the fish...

(There's something fishy in this parable, isn't it? )

Anyway, I initially wanted to add some explanatory details about each tech but it was too much work and I also thought that it would look a little ugly, so I left it unfinished, but if you would like clarifications, I can edit my list.

------------------
"In a time of universal deceit, telling the truth is a revolutionary act."
George Orwell

[Henrique Duarte]

I have little time but I´ll try to make my ideas clear, with as little chit-chat as possible :-)

Tech Model/structure
I agree we shouldn´t worry to much about the names we are giving things right now (call them theories/techs etc). As far as I am concerned the way to get the right simulation of effects of things like theories vs skills and the "language development" does not depend on changing the structure you presented. The tech model as is gives an excellent frame to create the necessary interdependencies between different techs/theories/skills (read this in terms of your tech levels). It is in the detail of this interdependencies and in the ways the player will be able to intervene that the simulation will succeed.
When I commented and agreed with F_Smith I meant that some care should exist in determining the relations between the different techs so that things like theoretical vs skill dynamics (basically level 1/2 vs lv 3 techs) are properly simulated. That was why I gave the examples of the roman joke about greeks thinking and romans making things, and the fact that much of the portuguese exploration in XV cent was made with a pragmatic/experience relied knowledge (high level 3 tech) without a solid theoretical development (vs comparatively low level 1/2 techs).
Anyway if I understood Richard well the different tech levels will interact continuously so, I will repeat, IMO it just depends on good number tweeking and not on a model restructuring.

Bottom line - IMO your on the right track, the model allows for a wonderfull flexibility, just wanted you to consider the theoretical vs skill discussion when the number crunching is made.

[Henrique Duarte]

Axi

Just read your post (after I wrote mine).
I suppose you said it all in what it relates to the theory vs skill discussion.

[Richard Bruns]

Henrique:

I understand you now. Unfortunately, we probably won't be able to model the ability to make things without knowing the underlying theory. It would create uneven gameplay if you could easily advance the Level 3 techs without having a good understanding of higher level techs. But don't forget that Level 1 techs are basically free if you manage your civ well.

Level 2 techs are a cross between knowledge and theory. They are not simply a less important Level 1 tech. They combine theory and practice and are an even split between the other two. So in your example, the Portugese would have decent Level 2 techs.

Axi:

IMO we are not the people to dictate the Level 3 techs. My goal is to create a good Level 1 and 2 framework for the other model people to attach stuff to. They know what Level 3 techs are needed for the game models they are making, so the Level 3 should belong to them, not us. For example, your tech tree did not include many of the Disease and Disaster techs that Toubabo_Koomi said were needed.

For these reasons, any quibbling we do about the Level 3 techs would be fairly pointless.

Toubabo_Koomi:

I haven't forgotten about your techs. It would help thigs a bit if you were to give an approximate date that you expect civs to be able to acquire the techs.

[F Smith]

** Warning **
(This post has been rated "B" for Boring by the people that rate those things.)
Idiot programmer ramblings most folks will not be interested in!!!

I'm sorry again. I'm sure this is 'too many cooks' syndrome on my part. This is ya'lls game, and I should just back off. I have a very programmer-oriented approach to analysis and design of programs and systems, that others are not always going to share.

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.

My greatest frustration is in that I just don't think I'm able to communicate, here. It is certainly my fault, my approach is too wonkish, my descriptions too object-oriented. This isn't a matter of words or titles -- it's one of basic object-oriented analysis. The interactions of the different parts of the system (the objects). This is beyond what ya'll have so far. Ya'll are in the 'analysis' phase right now. I just strongly believe that good analysis should start with identifying objects needed then mapping their behaviors, rather than later try and convert a system over into object-oriented code. Many, many times I have run into coding difficulties because Managers took this type of analytical approach with the clients. Its the norm in my company, where old Cobol architects and clients get together to design systems with procedural expectations and then expect us to implement them. We always throw out their analysis and start over again with the basic requirements, then floor them with results far beyond what they thought possible. Because the goal of OOA/OOD is to model reality smoothly and in a user-friendly manner. It *is* possible, and happens every day, as I'm sure many of you know.

I guess I should just drop it for good. I'm sorry, my fault. I just love to discuss this kind of thing. I think that's why I became a programmer.

It was just that I so urge you to reconsider several of ya'lls basic assumptions. My outside analysis of this approach is that -- it seems no less complicated than a 'connections' type system (it actually seems more complicated), 'StarCraft' style approaches do not break down upon scaling (I have tested this), and this system is in no way more conducive to a 'the player is the leader' interface (the UI makes them play similarly).

Perhaps later this year, ya'll can take a look at my other project and compare. (Mark, you probably remember the basic Real-time user interface I showed you last year?) The tech system is about half in place, and it is smooth, mindless and very, very playable. I will play ya'lls game, when finished. I'm sure it will be fun. I just sincerely hope you all will check out mine, too, to at least give it a chance.

You see, I am assuming that when the programmer really starts to do the actual 'design' on this system in object-oriented code, it's possible that the coder is going to begin to come across some of these issues. That person will need to define the objects that will interact, and give those object behaviors, as I'm sure ya'll know. I just always start my analysis with that in mind. And when ya'll get to that point, perhaps, whoever is doing the coding will see me as something other than a hopeless, idiot wonk (altho I am that too!). But a few tweaks to this system, with the OOD of the final program in mind, can gain ya'll power, flexibility and realism that you will not have if you don't consider program design.

Okay. Sermon over. The soap box is open. I will now go back into my cave until the next solar eclipse. I actually will be busy this weekend with real life, so I won't be bothering ya'll, distracting the discussion. Thanks for listening, and thanks for not throwing fruit.

Again, I'm sorry. Please don't hate me because I'm beautiful.

[Henrique Duarte]

F_Smith

As far as I am concerned please keep posting,
there is no improvement without confrontation of different ideas (just look at politics and the imposed death of ideologies nowdays)
By the way I still haven´t read something idotic from you, there may be difficulties in comunication but that works both for the receptor as well as the emissor :-)

Anyway I am not a programmer but I research natural systems (geologist)...read COMPLEX systems, and even admiting my ignorance (not as total as it may seem) in programming, I have found complex systems very difficult to reproduce in a "strict" object oriented approach. By object oriented approach I mean procceding to a "taxonomic" classification of what your dealing with keeping in mind your objective of characterizing the relations between the different objects.
I find frequently that frontiers between different objects are very hard to deffine - a flexible "classification" as to be made.
A subjective decision takes place when I have to put element A and B in different groups/objects despite their pontual similarities.
To acomplish reasonable results I rely on detailed characterization of the relations of as many singular identities as I can. I group those singular identities according to the "flexible" classification I made, interpretation/simulation is strongly helped by that classification but dependends allways on the researcher´s awareness the every "single element" is unique and can not be "boxed" and forgotten.

Trying to reproduce technological/scientific development in human societies is a REALLY COMPLEX problem.


What I am trying to say is that I saw in the tech models apparently "light" approach, in terms of hierarchy of "tech elements", that "flexible classification" that allows you to allways solve problems like - I know maths is more important than that it just doesn´t fit very well anywhere else.
As long as care is taken when chosing and characterizing techs and their relations you will probably work it out.

I will only say another small thing, I am all for the actual models approach (which by the way I don´t see how it can be similar to Master of Orion 2) but it will never work if the techs aren´t well chosen and their relations, no matter how complex, aren´t properly worked out.


Richard

minimum requirements for the invention of a new application vs minimum requirements for the use of an "imported" application.
This would allways be application related.

"your legions get that iberic sword (Gladius)(swordmaking 78%), your swordmaking is 50% allows you to understand the techniques behind the manufacture of the gladius, as the gladius as a base rating swordmaking 78%, the simple fact that you have the new sword will boost your swordmaking tech, to the minimum gladius requiremnt, in the following years as you already have acquired a "working prototype".

Also according to inter tech level relations, the acquisition of a new application will boost lower level techs (which is realistic)
inability to raise lower level techs to the point of sustainability of the new aplication may result in the "tech loss" consequence.

You could call it the "working prototype effect"

The "working prototype effect" should be per application.

[Henrique Duarte]

Richard

Just noticed some misspellings/confusions/bad grammar:

"your legions get that iberic sword (Gladius)(swordmaking 78%), your swordmaking is 50% allows you to understand the techniques behind the manufacture of the gladius, as the gladius as a base rating swordmaking 78%, the simple fact that you have the new sword will boost your swordmaking tech, to the minimum gladius requiremnt, in the following years as you already have acquired a "working prototype".

By boost see it as a bonus to the research in the specific application - not an instantaneous tech upgrade. Something similar to: make a large telescope - get x RP per turn on astronomy.

------------------
"Also according to inter tech level relations, the acquisition of a new application will boost lower level techs (which is realistic)
inability to raise lower level techs to the point of sustainability of the new aplication may result in the "tech loss" consequence."

I just saw that you call higher level techs to what I called lower level techs :-)

Also by tech loss I also mean loss of the prototype effect (just have to keep on beating those Celtiberians and getting a those swords)

[Henrique Duarte]

Can´t keep quiet today :-)

F_Smith

If you´re thinking about biology being a natural science and that they made up a taxonomic classification for all living beings just remember that it took them thousands of years to make up something decent (I don´t think we have that much time :-)) and that they are able to do direct observations of their object of study...something that any "historical" science isn´t able to do.

[axi]

HEY! LOOK AT THE LIST! IT HAS MUTATED!
(showing upwards, his hand trembling with fear )

I couldn't wait for you to ask me, could I?

Richard Bruns: If I have found 96 level 3 techs by myself, how many will they be if everyone inserts his own? Anyway, as I have already said, the tree is under constant revision. I believe that in order to define it's precise form in level 3 is to start mounting the applications.



------------------
"In a time of universal deceit, telling the truth is a revolutionary act."
George Orwell

[F_Smith]

Henrique:

P.S -- in Masters of Orion 2, the player selected his next research goal from one of 6 'fields of knowledge'. Each field of knowledge held 3 techs per 'level'. If your civ was 'creative', you got all 3 when you researched that level. If not, you only got one application. So the system had most of the things ya'll have -- fields of knowledge, general techs (the levels) and application techs. Ya'll have added a new intermediate detail level, but that will be transparent to the Player.

In Clash, it appears that the player will do very much the same thing he does in MoO2, unless I miss my guess. Altho the algorithm behind the scenes will be different, and there will be bonuses that cross over. But this will all be transparent to the player. And MoO2 was a fine game, don't get me wrong.

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, and in actual practical application have not turned out to be impossible to deal with on a large scale -- especially with 'managers' that allow a player to simply select goals.

Someone just *has* to prototype this, before you lock it in as *the* decision. How can you be sure how it will play if you haven't tried it? Altho, as I said, if it is just like MoO2, as my sample suggested, it will be fine. That was an okay game.

Not the best, but okay.

[F_Smith]

Henrique:

That's why I love this board -- it helps me to iron out and learn to talk about exactly these kinds of issues. This is going to help me in my job, big-time.

I had lunch with one of our System Architects, and I discussed this with him. He felt that your example of the taxonomic classifications was in fact the perfect illustration of OOA.

To whit:
To model all 'animal' life forms, one approach could be to do a large, taxonomic tree (much like the current 'tech' system being proposed). This tree will be huge, and never finished.

Or, you can do OOA --

Object: Creature

Instance Variables:
String name
Vector cells
Object DNA
Date date_of_birth

Public Methods:
birth()
growth()
death()
damage()

Then all you need to do is define the DNA class to handle all the different growths possible, and you have a simple, 2 object model that reflects every possible creature in the universe -- including their lifecycles. This is obviously more powerful than a taxonometric 'tree'.

This is the perfect example of the need for OOA. I thank you very much, this is an invaluable thing, for me.

Oh, and for the record, Alfred (the SysArchitect) feels that if ya'll don't do the OOA now, the person that codes this later will really wish you had . . . and sometimes the problems that arise can be fatal to the project.

[Henrique Duarte]

F_Smith

Taxonomy works very well for biology...then why is it such an headache to do a similarly good job in Paleontology?

Hint
The object of study of Biology and Paleontology is similar (albeit in the second the poor thingies are long gone :-))
Yet paleontology has to deal with a large time scale, which also means evolution, which also means much added complexity.

If you can explain to me (in as simple way as possible) how to simulate strong complexity, with plenty of unknown variables influencing the end product, plenty of interconnections, etc, through OOA, I will be happy. Because honestly I still do not see how OOA can do that efficiently (it feels too limiting for complex systems)

By the way, I also have a professional interest in this particular issue - modelation of complex systems (natural geological systems in particular)

Second thought that just occured to me. Did you think about why you could translate Taxon to OOA - You used DNA information - do you know how much theoretical work, Lab work, discussion underlies the definition of those strings of numbers? DNA reflects pretty complex information and IMO that is why your example works. By the way If you tried to fill in the variables in your objects you would find out that most DNA information of known life forms is totally unknown!

[axi]

LOOK! IT DID IT AGAIN!
(points upwards, his teeth clattering with fear )

Henrique and F_Smith:
This object oriented programming discussion has made me feel completely stupid, since I am unable to grasp a thing about it.
I suspect that Mr. Smith is implying that we are doing it all wrong, but still...

I am getting pretty angry with myself about this...


------------------
"In a time of universal deceit, telling the truth is a revolutionary act."
George Orwell