List of suggestions for various topics. #5
Description
Editing Game Files
So, obviously, you shouldn't have to edit files outside the game to progress through the game, since this is both not very intuitive, and will likely be perceived as a bug rather than a feature. I would also caution against using a scripting language as opposed to some visual drag and drop interface due to both the intimidation factor of editing code, and because the edit has to "compile" every time.
I would suggest something like my idea in the skype log, but since you clearly want a more in depth system of changes, you probably want more than just the ability to change the timeline for people and events, or layouts of the city. That said, there's not much more you CAN reasonably change, so I would focus on increasing the scope of the changes instead,
An interesting way to introduce this mechanic would be a forgery mission, in addition to the suggestions in my log.
People's Reactions to Change
In general, people shouldn't KNOW who's changing the timeline and/or map, and should respond with confusion, like "I swear this wasn't a dead end!". However, as the timeline diverges from the alpha timeline, people should start to realize what's happening, and become more angry.
As an aside, an important thing to consider should be how far you can change the world. Should you be able to change the layout of a building? A city level? Swap levels around? Can you change the alpha timeline to change people's reaction?
Overall, this mechanic needs some thought as to what you want to accomplish with it. As is, it's a very all of nothing sort of mechanic that's either super relevant or just irrelevant background doodads.
Faith
Since faith is a big part of the concept, there needs to be a way to interact with faith. The big question then is: do you let faith matter? I feel like the best option is to let some player choices decide whether or not faith helps. To this end, I propose a kind of RPG twist to the problem.
Suppose the player has two (or three) semi-hidden stats that the game tracks. One symbolizes the amount of faith the player puts in artifacts in general. This will probably include donations to shrines, the number of side objectives completed for the church and other conservative organizations, and in general how often you (try) to use the macguffins. This stat will, as it raises, increase the effect of the macguffins. Similarly, there should be a stat that keeps track of how much you use various tech in favor of the relics, how often you help the librarians and such. This will, instead of making the tech stronger, will improve the variety of tech, letting you unlock a better variety of gadgets.
However, there should be two caveats for this system to work well. First of all, faith should detract from science, and vice versa. Secondly, the effects of the faith relics should be subtle, like a small reduction to guard vision radius, whereas tech should have direct effects on the world, like a stun grenade, or a skeleton key. This way, having a lot of faith can make it harder to get the more tech oriented societies to confide in you and give you prototypes, whereas being Batman will reduce your faith in the shitty obsidian necklace, and make it's effect less pronounced.
An optional third stat to track would be something like a karma/general reputation stat, that simply changes dialog choices available. Doing generally paragon things will get you in the good graces of one faction, but do the wrong things and you'll find yourself in favour with the wrong crowd. This is independent of the actual reputations with each faction, and can bolster or hinder relations further.
Plot
Overall, I feel like this game would do better without a specific villain, and instead be more focused on an overall trend causing the ruin of Maya. Specifically, the fact that there is quite a bit of strife between the church/state and the technological sides of the city. Specifically, the game can have four possible ending classes, based on how you change the city.
The first two is a strict win for either side. Should you sponsor the librarians, or dishearten the church, people will lose faith in their macguffins, and generally throw them away, thinking them useless trash. Which would obviously be false. On the flip side, you can sponsor the religious leaders, and have people be way more pious. While this will not change relations between the factions, who don't really shun each other so much as they don't see why the other does what it does, there will simply not be as many librarians, and over time, knowledge will be lost instead of artifacts.
Another ending would be heightening the xenophobia between the houses, essentially creating a caste system. While this does specialize people further in the long run, it could also ruin what's left of the city's culture, if not history.
Lastly, an ending where nothing is changed at all. The city continues to decline a grain at a time, as it has for centuries, waiting for something, or someone, to upset the balance of power.
Of course, there can and should be many variations on each of these, but these are the obvious paths you can go down. Of course, there is no strict "Good" ending, since the fundamental goal of the game is to let the player find what they value more, which would be moot if you could have your cake and eat it too.
This needs a LOT more world building though, because there is very little flesh on this skeleton world.
Technical
If you plan to use the layout changing mechanic, it may be wise to use Unreal 4 instead of Unity. Case in point, Antichamber. The portals built in are VERY useful for this sort of thing.
Unless you have a specific library you absolutely need, you should NEVER license anything under GPL. This license has many problems with it that many people are willing to discuss at depth, but the main one you need to know is that it prevents you from effectively taking credit for the game, since a person can copy the source, change a single comment, and be in full right to claim that this is their work, based on yours. And while it is possible to audit this, no one will. It only gets worse if you want to monetize it, since someone can buy a copy and then resell any number of copies without having to send you any shekels.
Not likely the latter is relevant, but still important to know. I would suggest MIT or BSD, although ZLib is also popular.