
Originally Posted by
Packetdancer
Almost certainly not this. On a wild guess, memory/bandwidth considerations; smaller housing units are easier to persist to and from long-term storage, as a lower number of furnishings (and fewer customization switches to flip) means a smaller maximum size on the data blob being sent back and forth. Apartments staying within a smaller range almost assuredly means that they can instance more of them, meaning they can have higher availability.
Most likely, the housing instance servers have something like "an apartment takes X amount of bandwidth/memory, a small house takes 1.5X amount of bandwidth/memory, a medium takes 3X, and a large takes 6X" or whatever. (Numbers are entirely made up and just for illustrative purposes.) Meaning that for each large house, that same amount of resourcing could run 6 apartments, or whatever the real/accurate number would be.
Changing absolutely nothing else, if you just made apartment interiors the same size as large houses, it's likely we'd be able to have fewer of them 'active' at a time, and the 'you could not leave the area' error when there's no housing instance space available would be a lot more prevalent.
And while it's easy to say "just add more servers", it's also worth noting that the houses need to be saved to/from long-term storage when they're loaded in or unloaded. You need a centralized place for that, and it's quite possible to overload those systems when scaling up; witness the issue when the housing lottery was first introduced, where messages with the winning plot numbers got lost because of the sheer volume of lotteries. It was bad when it broke the lotteries. It would be so much worse if the things that got lost were the changes that people made to the houses they have; spending three hours decorating only to have it drop the changes on the floor would be heartbreaking and infuriating.
All of those problems can be solved from a technical standpoint, yes. But they're best solved at the initial time of server design, and the servers for this game were put together under the gun and incredibly hastily during the ARR relaunch; it's quite possible (arguably, even likely) that those issues cannot be solved with the current state of the back-end servers, so would require large amounts of redesign/rework.
I would actually love for there to be fully instanced housing, because I'd love for everyone to have housing. The reason I personally look for alternative solutions -- decoupling FC workshop or gardening from house ownership, moving FCs onto instanced 'airships' (because there are fewer FCs than players, and if you require a certain minimum rank and membership level, you put a soft cap on the number of instances needed) and out of wards (to free up space for individuals to get the housing plots), etc. -- is just because I suspect fully instanced housing is not feasible under the current server structure.
And while I would not be surprised if the devs, e.g. CBU3, would actually like to redesign the servers such that they could support it -- because let's be honest, anyone supporting a ten-year-old codebase probably wants to redesign it anyway -- I doubt that the publisher, e.g. SQEX, would sign off on it.
(Yes, it's easy to say that SQEX gets a lot of money off of FFXIV and that the money should go back into it, and that getting a team to rearchitect the servers to something more modern and less hastily-slapped-together during the ARR reboot would be a wise investment. While the statement is true, that does not change the fact that it isn't how video game publishing works... at least not at any game publisher I dealt with, in my personal experience. And while none of those were SQEX specifically, I certainly suspect the same is true here.)