Wow! My sincere greetings to a colleague
I'm working on Odyssey planetary settlements mapping project since the very first day of alpha (oh, for more than a year already). However after initial data acquisition and some consideration about underlying "procedural" game mechanics involved I've decided that the "screenshot-based map" approach is not sufficient in that case - in contrast with the "legacy" (static) Horizons settlements.
Reasons:
1) Building interiors - not really appropriate to represent them with screenshots
2) While legacy (horizons) settlements are just a set of absolutely static models, those in Odyssey are quite different - we have 28 layout generation templates that produce results with a degree of variation (both of procedural/persistent nature using template-supplied rules + final instance-dependent RNG touch in the limits of the result of previous operation, and all of that to a degree affected by other factors - security/state...).
As far as I understand technically only general layout - walls - is truly static inside the given generation template.
Most (if not all) interactive objects - ports/switches/lockers/med-energy boxes and points/explosive barrels/etc - at least technically could be placed differently in the resulted layout (or just simply absent/present) based on the same template. I could only assume that those objects can have a limited set of possible locations (and for some functionally "important" objects those "sets" can degenerate to only one single possibility in case of particular templates). So not only data ports (DPs) can appear/disappear/change placement from one settlement to another having the same template, but even Settlements Alarms Console (SAC) switch has been documented as been placed differently - different rooms but of the same area - in case of some particular layouts. De facto this could be (or at least was) illustrated with at some period earlier frequent generation bug when a 2-cell sized prop is superimposed with different 1-cell size prop on a 2-cell size "placeholder".
Also I can suppose that even purely "decorative" content is not always "nailed" for a given template (as it was in my initial assumption). A have a case that needs rechecking to be sure, but once I've stumbled that the entire content of the room have been noticeably changed based on the same generation template - a hangar with boxes/lockers replaced with the same hangar but with an SRV wreck.
So I've decide to take a way more "hard" way to draw maps of the Odyssey settlements - fully reassembling them using vector graphics (
an example of my work could be found in this post)
As I've said above, static are only layout templates, but results generated upon them can vary. Since alpha I've already visited few hundreds of different settlements, sometimes spending not hours, but days to record/collect matches and differences with "sister" layouts. Contents of those lockers is another story - it's an instance dependent product of RNG that in case is limited by coefficients generated on the base of the template and also affected by other parameters (like security/state).
so definitely
You know, at the present moment I do have at least one documented location for each of 27 different/"unique" layout templates, including:
5 Agricultural
5 Extraction
4 High-Tech
5 Industrial
5 Military
3 Tourism
27 in total, and the last - 28th - remains still undiscovered by me=) And I'm still looking for it
Yeah, I suppose that original naming is "something + economy + template_number_inside_given_economy" (I do remember bug reports referring that approach). Now we have size (only 3 gradations) and security (I think that the latter does not affect the template itself, only imposed on generation "rules")
Heh, good question what do really depends on security/threat level. For sure it defines the rank of NPC security guards (except special ones)