Bay 12 Games Forum

Please login or register.

Login with username, password and session length
Advanced search  

Author Topic: Near term development suggestion: world generation overhaul  (Read 384 times)


  • Bay Watcher
    • View Profile
Near term development suggestion: world generation overhaul
« on: December 13, 2018, 04:17:19 am »

I can't exactly knock that world activation in 40.0 from the archaic form of a singular world of left to right and even the jump from 32 bit to 64 bit weren't momentous leaps for the game, but almost as readily as the upcoming 50. editions, with the content coming thick and fast it's sort of been revealed that without nessecary updates, the clean-ness of the new C civilisation viewer really makes our currently world-generation look out-dated.

Improved world generation screen

* Better handling of event instances: It is no secret that world-generation functions significantly differently, and this isn't exactly nessecary for it to be this way as it will generate different outcomes as the considerations added to the raiding arc are not applicable. World generation is particularly important to creating a setting and hence would need the consistency of regular play world generation in order to create a fluidity of the in-world narrative for events.

* In many ways currently, in order to have a true and intentional depiction of how civilisations are fighting and megabeast duelling with heroes and settled sites across the land, you have to start at date 0 and play for a significant amount of time post-generation, which in real life terms can last weeks in order to make significant progress over time, which with the difficulty curves of the game can be very boring for turtling or difficult requiring restarts and save-recoveries.

* In perspective, a hundred years of in-game play and background w.g can be anything between a 100 hours on good FPS, or 300 approximate hours on poor FPS in larger fortresses which would equal a approximate delivery date at 125 'short' worldgeneration setting at comparison over a real life year or longer with only 3 or so average hours availible to play DF in a day if not trying other activities. Hence its as important as ever to provide the most concise world generation as to ensure the best player experience.

  • Onhand information and pre-legends accessibility: A big caveat of the world generation screen is a distinct lack of information, you can observe the basic state of every site as to its location and much of the information is inferred later, a direct port of the level of information from the C civilisation screen in fortress mode here would help, including especially the site legend and a active readout of important information like approximate population.

Improved legends screen

Interactivity & UI accessibility: The legends screen is often outsourced to statistics by third party viewers which offer more of the kind of nitty gritty information about the overall and intricate views some player's want in order to understand their world, whereas basic site information and links can be more easily inferred through interaction which is a big thing i feel has been neglected to take advantage of in the legends viewer.

Spoiler (click to show/hide)

* The enclosed image from a interactive third party viewer in which specifically i've outlined sites to select sites, without meaning to strictly look at global charts through other functions of the viewer, i can interact with any highlighted site and instantly pull up their personal event file like finding them on the list in existing legends mode or work seperately through name and correlate their position on the map and events at a glance. This is all using data already exported via XML but to simply make something similarly as standard would significantly improve the functionality.

Here's the legend viewer's forum thread and Github release source as to raise Toady's awareness of it.


With a mixture of these points, I believe a entire arc portion should be slotted in first to address these due to the ambiguent uncertainty over the 'big wait' for the magic arc delaying development of the game, which seems to be a shared worry with other people in the community.


  • Bay Watcher
    • View Profile
Re: Near term development suggestion: world generation overhaul
« Reply #1 on: December 13, 2018, 12:32:12 pm »

All of these things are on the huge "to-do" list.  I think the issue at the present time is that the myth and magic arc is going to fundamentally alter world generation.  So improvements to world-gen are going to come afterward.

It is currently undecided whether the various race's creation myths are going to be literal or interpretive.  Also, the role of magic as well as direct interventions by gods(?) will shake up world-gen events.  So I think that indirectly the myth and magic will improve many of the things you outline.

As far as the Legends Mode UI, it is low priority only because the general view is that any significant time spent on the UI is time not spent fixing bugs and adding features to the game.  In other words, 'it would be great but we want other stuff more' is the attitude.

That's my take on the current situation.

That being said I too really want world gen and post world gen events improved!  The past couple months I've been playing in a single world.  I've built four different forts between the years 504 and 533.  I've also done some adventure mode between each fort.  Between each one I export-legends and take a look at what has changed.  Having older versions of the XML to refer to is useful, as some changes are retro-active and alter some of the historical info.

I would say that warfare is the biggest oddity.  Many instances of one person attacking a whole town.  It seems like is one member of a civ gets in a tavern brawl with another it gets logged as a conflict and attributed to a war.  That's the only explanation I can think of. 

The upcoming villains and plots should add more volition to the actions of important figures in the world.  Honestly I'm going to reserve judgement until that is released.  I'm guessing it will affect world-gen quite a bit.


  • Bay Watcher
    • View Profile
Re: Near term development suggestion: world generation overhaul
« Reply #2 on: December 13, 2018, 05:19:37 pm »

I also think we'd need to see the big wait overhauls come by before worrying too much about DF's problems with exposition(as in, explaining you about the world) and abstraction. One of the things mentioned in the gdc talk was the notion of giving the player a version of the creation myth somewhere during or after worldgen to let them settle in. Having the ability to discover and name regions would help with this as well.

As well, in the infidelity devlog there was mentions of improving the historical figure legends page. There's been talk about getting a hyper-linked legends UI as well, like the preview mythgen has.

Honestly, having done the same as Sketches, I feel that one of the problems I am having is that worlds and civs are still a bit too similar right now, so I am super stoked for the myth arc to shake up the former and the property and customs arc to shake up the latter. (this is especially bad with vanilla elves. Dwarves have different artifacts, different mountains, goblins have different demon masters, humans have different values, elves have... different animals???)
Stonesense Grim Dark 0.2 Alternate detailed and darker tiles for stonesense. Now with all ores!


  • Bay Watcher
    • View Profile
Re: Near term development suggestion: world generation overhaul
« Reply #3 on: December 13, 2018, 09:14:02 pm »

That and the ability to shape living trees into objects, which we don't even get to watch happen right now. Luckily, I think the myth arc will spice them up a bit with their own unique forms of magic and such.
It's actually pretty terrifying to think about having all of your fat melt off into grease because you started sweating too much.