I guess the ruler approach is realistic while communal decisions would be hard to develop, because NPCs would have to make decisions and have opinions on a dynamic scenario. At first glance, very dificult, but I am working on a decision making algorithm that may allow for this down the road. We will see . . .



Would you care to elaborate on this?



Supposing you start with 3-5 NPCs, it is not enough to reach a commercial seafaring stage. But building a raft is quite realistic.
Just thinking aloud: the player could have/find a map/note that says "To reach the continent, follow the north star" and he has to row/stir/drive the raft in the right direction and reach civilization within a certain time limit or die at sea. This would be fun.





Yes, going mainland to recruit/hire more people to the community would be quite immersive. Mainly during the initial stages when the community is small and word has not spread yet. On later stages, people may start to show up voluntarily, but we still could have some quests to find some special people (like a master craftsman or such).

All recruitable NPCs would have to be added by the mod (to avoid conflict with other mods), which means the player would have to have clues that lead him to the right ones, which means fairly elaborated quests.

At some point down the road I will start to ask for help with some aspects of the mod. I suppose the implementation of this kind of quests will pretty much depends on how much help I will get (*crosses fingers* *knocks on wood*).



Yes, they are. Very creative. Thanks

A thought as far as population:
Two tiered recruitment. You have generic npcs that you can acquire several ways (notices in inns, buy from slavers, ex-prisoners, etc.) and skilled npcs with quests related.
The skilled npcs are your money-maker, here. There can be simple things like the farming family mentioned above: get them to like you, supply some scratch and tools to get started, wait, and you have a few extra skilled farmers that supply a bonus to food production. Or, on the complex side, an apprentice smith is running around town, looking for someone to help him save his kidnapped master. Once you find and free the master, he tells his apprentice that he was poisoned or something and leaves a cryptic map and key. Following the map leads to an ornate chest full of design manuals that show the apprentice how to make the legendary armor/weapons/whatever the master was known for. He then joins you and can produce said legendary item or just superior quality items in the armory (or if you want to add some customization, choose multiple places he can be assigned: armory, tool shop for manufacturing bonus, for farm for production bonus for better tools and such).

This is quite doable with the planed engine, as I already intend to include NPC skill levels. So we could have from unskilled NPCs to masters and everything in between.

Thanks.

And for those that have not noticed, I published the http://www.gamesas.com/bgsforums/index.php?showtopic=1060697 I developed to control NPCs in this mod.

When you say NPC skill levels, will they be static or progressive? While it does make sense for, say, a smith to eventually become a master of his trade, I wouldn't be surprised if this unbalanced your economic engine in some way.

I am planning it to be progressive (but very slowly)
Also, one thing that is on my list of possible features is enhancing the skills (or production rate) by other means, like the design manuals you suggested, or a unique tool. Just a l few thinks like that, here and there, as I don't want to develop a Tech Tree like the Civilization-style games.

Balancing the mod behavior over time will be a major challenge, anyway. I may have to drop a few neat ideas on this account.

I was reading this http://www.gamesas.com/bgsforums/index.php?showtopic=1060142&st=0 thread and a crazy idea just occurred to me:

How about making The Evolving Society as a Total Conversion mod?

Meaning, when you activate the mod, you start a new game. No connection with the original game whatsoever. You and a couple of others wake up in a beach as survivors of a shipwreck. Then you spend your time building up the community until you reach max development or max failure. Something between 2 to 4 hours of game play. Replayable - different evolution each time you play.

Actually, I don't see any advantage and I don't think I like the idea that much, but I would like to hear your opinion.

The only thing that attracts me is that it has not been done before (or maybe it has . . . I don't know)

If its implementation doesn't require a lot of work, why not ? Such a gameplay mode will certainly get me to play the game

Actually, it might mean even mean less work. At least for the second prototype.

This mod 'dependency' on Cyrodill is mostly:
- The initial lore to get the player to the island.
- Quests which main objective would be to boost some aspect of the production process with some resources that would not be reasonable to find in the island (technical manuals, master craftsmen, advanced tools and such).

I could leave these quests to be done last and, for the initial beta version, pretend there is Bermuda Triangle nearby and, from time to time, some crates/survivors will show up at the beach.

This would allow for fully testing the game mechanics before spending too much time on elaborating the details (e.g. Quests)
It would also tap the idea of a Total Conversion (in a sense that the player starts the game in the island and cannot go to Cyrodill)

I have two major concerns with the total conversion idea: Where would immigrants come from and who would you do business with?
I see a couple of solutions ranging from simple to very complex. Creating a "shadow" land that really only exists in name and statistic form would solve both problems, but eliminates possibilities of quests for advanced tools and workers and such. Another settlement on the landmass (needing to be found by the player or a scout, naturally) would also work and be more interesting as well as getting those quests back. This could also give you a way to beta test your mechanics before introducing the more complex aspects. My preferred solution would be a separate land mass with a water divide between you. This a) solves all issues I have listed, b ) gives potentially more time for mechanics testing since players need to develop a navy before exploring the new area, c) opens up great possibilities for creative thinking (the second area could have a very different culture to point of just being interesting or requiring some diplomatic dealing to maintain good relations, depending on how ambitious you get). A combination of another settlement and another landmass would also work without too much issue, I think. Only problem is the amount of work this could create. I'm going to guess it'd be approximately a whole lot.
I can't code or build textures or anything, but if you do decide to implement the whole foreign culture thing, I think I could come up with a lot of good stuff.

That sounds interesting...start from scratch as shipwreck survivors.

Yes, those are legitimate concerns.

For the first version, a shadow land or shipwreck survivors will do.

But, for further versions, a whole new landmass, with new cities, population and culture, seems impractical to me: too much work with very little to gain. That being the price of a total conversion, better drop the idea and go get immigrants and goods from Cyrodiil, which is already made. Although, as you say, it would open up some possibilities . . . maybe as an AddOn further down the road!!?

An existing settlement in the island kind of goes against the initial idea of a wild island, but may be necessary, depending how things go.


On another token, as a status report, I have two good news:

"Habemus Landmass" - I accepted the offer to use an existing, fully developed worldspace. It is a beautiful one with a lot of variety, that fits perfectly with the concept of this mod.

I finished the code that will be used to control the NPCs on the island. I will, now, resume working on the buildings positioning and replacement system, which, IIRC, is the last "building block" before putting everything together and, actually, start the development of the mod itself.

Qquix, remember you can't change Walk paths that creates some problems if you want to add the posibility to build a house
I don't know if you was assure of this but its pretty essential if you want to add the feature

Btw this mod is really a great idea! And you're really doing a great job

Yes, pathgrids are THE major technical problem for this mod, as NPCs wouldn't know about new buildings placed on their way and would try to walk thru them.
Actually this is a VITAL problem, because if not solved, the whole mod concept goes down the drain.

I've thought of a few possible workarounds, but I am not sure of any of them:

1- One (insane) solution would be linking every path node to nearby rocks/trees and disable them (with DisableLinkedPathPoints) when I disable the rocks/trees to clear the area for the buildings.

2- Pre-determine the city/village locations and pre-determine their street 'design', so you can only place building on pre-determined areas (where there are no path nodes). This is the solution I like the least, because I would like to keep the city location completely flexible.

3- Do not create any path grid at all and dynamically control the NPC movements, considering the existing buildings. Actually, this potential solution (the need to control the NPCs' every breath) was what led me to the development of http://www.gamesas.com/bgsforums/index.php?showtopic=1060697 (although it took a life of its own, after I started it).

4- (in addition to #3) Requiring the player to, somehow, determine where the streets are, maybe by lying some pebbles, maybe by walking around while a script monitors his movements, or any other way of dynamically simulating a path grid, so to speak. (I really don't know . . . I've just thought about this while writing the previous one)

As always, ideas and discussion are welcome.

How about using path grids with the above ? Let the actors use normal path finding and let the script do periodic checks to prevent collisions and such unwanted behavior. Should require lesser processing power than #3 alone.

My experience with trying to detect NPCs colliding with statics did not get good results. When trying to detect the walk-in-place scenario, by the time the scripts figures out that the NPC is walking in place, it is too late - it already looks bad in game. If you try to make it more sensitive, you may trigger the reaction in other situation when you should not. Besides, I am not sure what I could do to fix the situation, once detected.

Or, maybe, you have something in mind and I missed the point !??


I was thinking that if, somehow, I could determine where the streets cross, I could dynamically place markers at the crossings and use AI packages to make the NPCs travel to the first marker. When he gets close enough, switching to a new AI package directing him to the next marker, until it gets to the destination. This could be done with or without path grids, I guess.

Easier said than done, of course. I have no idea how to determine and store the street design and, worse, how to determine the order of the markers to get the NPC from point A to point B.

Thinking about it, I could force the buildings to only be in multiples of 90 degree Z angles, so the 'streets' would always be north-south or east-west. The math would be easier to figure out.

One thing that may help in this whole thing is that I will know the dimensions of the buildings footprint (this is my next project).

Adding some simple building guidelines might help that. If it's just a few restrictions (certain degree rotations, for example: 15, 30, 45, 60, 90, etc.) then it's still very much within normal city planning and will only inhibit really bizarre designs.

My thing against total conversion would be... well I only have the one copy of oblivion on my system and I haven't actually finished the main game. I mean it'd be cool, but I'm sure it'd be a pain to switch between the two. I mean I know there is a way to install two versions of oblivion, it just seems like a waste of space for only one TCM.

Unless there is a way to let you chose what game to start from one install like the alt-start mods that some how allow you to do the vanilla openning or one of the others.

But that could cause conflicts ... the seperate landmass /worldspace idea sounds nice.

Yes, I guess I will have to reach a compromise at some point down the road.
But I think that this problem will be more clear / better defined when I get a couple of cities dynamically built. Anyway, I anticipate a huge problem.


I suppose it would be just a matter of managing your saved games.
You save your present game before you activate a total conversion mod.
When you enter the game after installing a TC mod, you are teleported to the mod worldspace and play, save and load normally.
When done, or need a break, you deactivate the TC mod and just load your original save and you are back in business.

But I am no expert in the area and things may not be that simple.

Talk to the guys who made RTS (Real Time Settler) for FO3 i think they figured it out.

What, the path grid problem? FO3 doesn't use path grids, it uses a navmesh.

I was preparing to post the question in their thread.
If FO3 has a different engine design, any solution they may have would not apply to Oblivion. Did I understand it right?

Exactly. Unless Bethesda adds navmeshes to the Oblivion engine and lets us mark interior cells and worldspaces with a "uses navmesh" flag when we want them.

The solution in FO3 - and I'm guessing here based on the GECK wiki entry for navmeshes - is simply to drop a very, very heavy havok'd object wherever you want the navmesh "broken". Actors will simply navigate around the area.

I can imagine how it is done . . .

Hmmm . . . by positioning each building inside its own subspace I might get a similar effect . . . I wonder how Oblivion engine handles dynamically moved subspaces . . . time for some testings . . .

[EDIT] Did not work. Subspaces do not prevent NPCs from crossing them. Too bad. I had high hopes on this one.

Lol my bad im not a modder. Thought they both used the same stuff.

I've been working on things leading to this mod almost every day and I suppose I could call this phase a milestone.

So . . . progress report time:

WHAT WE ALREADY HAVE
Three very important 'building blocks' completed and ready:

1. http://www.gamesas.com/bgsforums/index.php?showtopic=1060697 - A system to have total control of the NPCs in the island. They may interact, if necessary, and they will keep working (and producing) even if not loaded. Besides, the system will allow for many secondary, but cool, possibilities, like guards dynamically choosing their armor.

2. The Construction System - the 'mechanics' for creating, replacing and removing buildings is ready. With this system, the player can build her city from ground up, starting with a few, cheap houses to a thriving city as big as she can afford with available resources. It will be possible to add walls to the city, trees to the parks and sidewalks and lights to the streets. The system is in Beta test and will be released as a separate mod soon.

3. The Island - I accepted the offer to use the island created by the 'NPC with Jobs' team. It is a most beautiful island that has all we could wish. Still pending authorization to do small adjustments to the landscape, if this proves necessary for some reason along the development.


NEXT STEPS:
1. The Production System - definition of the raw materials, the superior materials that can be created by combining them and their use to create new buildings. There was a production system in the prototype, but OBSE has evolved so much since, that I plan to start from scratch. And this system will also depend on. . .

2. The Lore - There are A LOT of ideas and suggestions proposed in the three mod forums. I, now, plan to classify/merge/organize them all and present a summary of choices to be discussed in the forums. So, if you still have some idea or suggestion to present, this is the time.

3. Further down the road:
- 3a. The relationship of Buildings-NPCs-Productivity (as in which buildings will make which NPCs happier and more productive and which NPCs less satisfied with the player's administration)
- 3b. Look for a team to put everything together.

MAJOR CONCERNS

I have the solution for the LOD problem: the city will be visible from anywhere. This is out of the way, now.

So the major problem left is how to make NPCs go from one place to another without bumping into buildings the player has placed in their way. Luckily, scruggsy is willing to look for a way to add functionalities to OBSE to dynamically change the pathgrid. If this fails for any reason, I either (1) have to create code to mimic the engine path finding (very complex and time consuming) or (2) pre-determine the city and villages locations and restrict the player to place buildings only in predefined 'lots' (which is against the very concept of the mod)

A minor concern, for which I don't expect a solution and we will have to live with, is the land texture. If you place your city where the land is green, the streets will stay green.

And, of course, the big question: what size of mainframe we will need for playing this mod, when everything is running at the same time. LOL.


Ok, then . . . as Don Quixote used to sing: ". . . And the wild winds of fortune will carry me onward, whithersoever they blow. . . "

[This same text has been posted in the three mod forums: http://www.gamesas.com/bgsforums/index.php?showtopic=915021&st=0 - http://thenexusforums.com/index.php?showtopic=83048&st=160 - http://thewormhole.nfshost.com/forum/index.php/topic,1298.0.html ]