My case for relocations being a bit more de rigeur remains pretty much the same: throughout history, examples of cities being 'razed' and thus swept out of existence permanently isn't nearly as common as a city simply paling away compared to other nearby cities. Indeed a lot of cities that were actually razed to the ground - Carthage by the Romans, Samarkand and other cities by the Mongols - bounced back very quickly.
I agree, from the examples I can think of what is more likely to cause a razed city to be displaced if either an occupier or the previous inhabitants deliberately decide to rebuild at a different location, which then eclipses the previous center for political or defensive reasons, part of which is why I think "relocation" encapsulates this well as a concept, although it covers more than "city A has been conquered and city B has been established as replacement".
On the other hand, cities like Babylon stopped being inhabited when river paths no longer irrigated the city and desertification in Mesopotamia increased. Of course there are different approaches to how far you can go - for example, I have York become Leeds when England reaches the Industrial Revolution, which is an accurate reflection of York's dwindling importance compared to the cities that were closer to the mines in that era. But of course York is very much still inhabited - I live here!
Yes, if we want no Leeds in the middle ages but also some games where Leeds can appear at all, then we need a relocation like this. And I think I agree with this goal. The question is how we communicate that to the user, and if we give human players an option. At this point I just mostly want to bring it to everyone's attention that there is a significant difference between relocations and mere hypernames when it comes to player experience.
Currently, the only times a city will be renamed are: when it flips to (or is conquered by) a new civ, when a new era is reached, when a civ becomes Communist (almost all Russia, a few Soviet sphere of control cities (like Ulaanbaatar), one or two Chinese cities and a slightly different transliteration of Korea's core cities) and finally, and least used currently, when a city adopts a new religion - currently only used for the notorious Buddhapest and also the city of Gaya/Bodh Gaya, which can't currently be founded in the game as long as Patna exists, which I'd guess is every single game the human isn't playing as India. I think an interface announcement would be great, of course.
Speaking of, are there more candidates for religion based city renaming? How about changing from transcribed Nahuatl names to proper Spanish names from example, instead of Nahuatl -> Spanish directly when Spain conquers an Aztec city?
This is implying a full relocation to a different tile, I take it? I'd support this in some cases, as with pesky old Babylon. I was thinking Persia could relocate it to Tisfun/Ctesiphon under some circumstances, which feeds perfectly into Baghdad. Of course this becomes thornier with the Carthage example, where Tunis occupies the same tile.
No, when I say relocation I always mean keeping the ingame city on the same tile, but using the name of another city that occupies the same tile but is a distinctly separate city (although it may be incorporated into the other at a later date, as with Carthage/Tunis). I don't think it's possible to move the city to a completely different tile, there are too many exceptions and edge cases to account for.
That's fair, although currently the vast majority of relocations do come from conquests and collapses - there are a handful from era, as with the British Industrial Revolution example before, but they're far less frequent. It maybe just happens more than you'd expect - India does change hands a lot, after all! India in particular is a really difficult one because with very few exceptions (really Delhi, Patna and Karachi if you consider it to be a continuation of Debal) there aren't really any cities from classical or even medieval India that are still significant in the modern era. Seeing Mumbai and Kolkata founded that early in particular makes very little sense, especially when cities like Sopara and Tamralipta were among the most important ports in the world in their era. I do like the idea of cities being able to pick from a select list of cities for certain tiles - there were times when it was agonising to leave certain cities out.
One thing I will say is that while it looks very chaotic in an AI game, where I've set the whole world to be visible and am watching every single little detail, this won't be the case for the human player. For a start you have a far more limited view of the world - as far as I can tell, most people who play tend to see their own empire and its immediate surroundings and not waste hammers on exploring too much unless its to colonise. So naturally you'll see less of the AI's name change machinations - if you're playing as Portugal, you might see Kandy or Anuradhapura or Polonnaruwa or Galle when you reach Sri Lanka, but you're not going to see it cycle through all of them before you take it (and establish Colombo, naturally). Secondly, if you're the human player... well, you'll hopefully not be haemorrhaging cities to independents and other civs on a regular basis, so again it shouldn't be too confusing! Unless you lose a city, you'll only ever get the era-related renames, which are mostly hypernames. Again, could still be confusing the first time if you're not expecting, say, Oslo to change its name to Christiania for a few centuries before reverting, but for the most part I think it's manageable.
Yeah, that's sort of my line of argument too. Relocations are not
functionally different from hypernames because in each case the player is confronted with a new name for a city on the same tile without a proper announcement of the new name (unlike renaming on conquest), but "allowing" relocations as well simply increases the number of times where this is encountered by the player, both for own and foreign cities. It's not as much of a concern to me if it only happens once per cities, but some of the examples (like in India, as you mention) the screenshot descriptions seem to suggest some cities relocate multiple times during the game. I acknowledge that it's less of a problem when it is happening for the AI, which is why I mentioned different treatment for human and AI cities (or at least a player choice), but I also don't think it's not a problem at all if it is happening excessively for even the AI.
I share your goal that we want to see classical and medieval and modern cities in places like India, and the no relocations policy would not accomplish that goal given the region's history (and we're either stuck with classical cities forever or anachronistic modern cities like now). If it's true that change of owner is the principal trigger then that's fine for me. After all, even in the game we have India, the Mughals, English (or other European) colonies, and back to India as the intended chain of owners, with Tamils and the occasional collapse (and therefore change to independents) as possible triggers already, which should provide ample opportunity to cycle names throughout the game. But on the other hand if there's let's say a human controlled India that never collapses or loses cities it would be kind of cool if the classical cities remain in place because the historical events that led to their decline were averted.
I sense from some of your screenshots that the desire is to showcase all the names the prominent city on a tile had throughout history, and it's cool to exhibit as many possible names as possible, but I would prefer a system that is less rigid and more responsive to what is happening in the game, but leaves the possibilties open for multiple names to be selected so they are eventually all seen over multiple games. You already mentioned that some names can only be seen under very specific circumstances and I think it's kind of cool that there are some rarely seen names that you can be excited about encountering. Definitely makes the exploration part of the game more interesting than just finding the same old pre-ordained cities every game.
Very much in favour of point 6 and tentatively in favour of point 5, though I'm aware it may create a few oddities - as an example, we might see a Tibetan rendering of Constantinople in a run where the Romans and Byzantines never touched Byzantion.
Yeah, if I do this it will be implemented so the game is aware of the renaming "order", e.g. Byzantion -> Konstantinoupolis, and it will only check back into the past but never into the "future" for renamings that have never happened. So if you don't have a name for Konstantinoupolis you can fall back on Byzantion but not vice versa. Also, this could allow us to tack some translations that are kind of irrespective of the current hypername (i.e. the "Tsargrad" chain - although that may be even better as its own hypername?) to Byzantion and then it doesn't need to be repeated for the Constantinople and Istanbul alternatives.
That's interesting, I'll have to do a little experimenting! Explains why Carthage wouldn't found two Qart-Hadashts. Only real problem with Edo is that both variants might take different rename from the same language - English, for example, as you'd want them to rename the African one Benin City (and keep the Japanese one as Edo). It very well could still be possible though, I'll have a dive into that when I get the chance. Fortunately I think it only really applies to Edo and a few Spanish cities, though there again you'd expect Spain to found both Santiago de Compostela and Santiago de Chile.
It also explains why Rhye chose to be wordy with all the Alexandrias. Another one I thought about was York, i.e. York (England) and York (Ontario), modern Toronto. It's not as unlikely to happen as I thought (which you could argue for Edo).
It's true that changing the DLL so that it accepts identical city names is possible, but the question is also if that is a
good idea. Suddenly, notification and interface information isn't as meaningful anymore because you don't know which city the game is referring to anymore. For some, we have real world disambiguative names like in Spanish, and I could even imagine writing code that only adds the long form if the name is otherwise ambiguous. But for other names the best I can think of is programmatically adding the region name to the city if the name is a duplicate, e.g. York (Ontario) and York (Britain). I don't think we ever have the same name twice in the same region. I am using the larger set of regions from the big map here, by the way.
I also discovered another problem with the current code. If two different cities have an identical name (even if not in the same language), but different identifiers (as they should), the game cannot really deal with it. For example, the Phoenician name for Malaga in Spain is Malaka, but that is also some translation of Bandar Melaka (didn't check which language). So it's impossible for the code to determine which city it actually is when encountering "Malaka", leading to mistranslation in one of these cases. And the code isn't easy to change because we only get the name string, not the full city object, so it's not even possible to go via tile or region instead. That's another argument why my planned transition to unique identifiers for cities on one single city name map is a good idea.
