I understand that some players run multiple accounts. Well, how difficult would it be to have, lets say, 2 accounts run on one client? You can pay CCP for the service just like you do with multiple training. Just swap pilots by clicking on a second portrait made available once you pay for it. The interface would change accordingly just like you were running two separate clients. It eliminates the mess of two separate interfaces by making them swap-able on just one.
This would run great on the coming 64bit client as well given all the extra ram access.
Based on past discussion on being able to swap between characters on the same account without logging out, this would probably be impacted by the same constraint that has resulted in CCP not reworking the codebase up to now.
Once a connection has been made, both the server code and client code make a lot of assumptions that the characterID will not change, and it’s very far from trivial to change that (otherwise we probably would have done that a long time ago)
That same issue will exist, if you wanted to switch between two characters on different accounts. The characterID would need to change for that to happen.
Unfortunately then, the 64 bit client alone couldn’t resolve that, since the server side code would need to be rewritten too.
It’s not that CCP don’t want to do it, but that because of the amount of work involved, it just doesn’t receive any priority.
And I think that is where CCP fails as a business. It seems to invest time and money into the wrong things.
The ease of a player being able to play alt characters from one interface is a money maker simply because it makes the feature available to use to less tech savvy (or people like me who just dont want to bother with the technical aspects of running 2+ clients). Their revenue would increase IMO, as more people would opt to pay for the additional character slot.
IMO It’s a good suggestion. I think CCP would respond differently if there was no technical issue.
But you’re almost certainly underestimating the cost of the necessary technical changes by at least an order of magnitude. Changing an essential low-level UUID from “exactly one” to “many” in old code is just one of those things that’s usually quite difficult and a lot of work.