ok thanks. I re-read the post you made and it seems where you said “The answer that ends up being arrived at for this question will really depend on how the technologies we’re working on end up functioning” was where I misunderstood. I believe you were referring to functioning as in compared to how traditional servers have traditionally functioned, not suggesting you are working on something that might not work at all. :slightly_smiling_face:
and the question was relating to server up time and how long. I know eve online restarts their server cluster every 23 hours and used to have a downtime of roughly one hour but now the downtime is about 10 minutes.
isn’t eve online a good model to learn from in terms of server? They been doing fairly large universe, thousands of players and all in the same universe for quite some time. (granted they don’t have ship interiors and planetside. and they do have some load screens like when you dock and it has that ‘session change’ timer as it negotiates loading the station you dock at.
oh and jumping from one star system to the other is a load screen too, hidden by the wait for the jump through the stargate. but still must be something there to learn from right? :slightly_smiling_face: they still get thousands of players in one battle even if it runs slow and terrible.
I feel you guys took the real hard route lol in terms of immersion. elite dangerous load screens are behind every jump between systems. gives time for each system you arrive in to load while in the jump tunnel :slightly_smiling_face: but its still fancy, I guess that’s why I prefer sc, everything is real time and full immersion.
thanks for your work.
Every MMO that I’ve every played/worked on has to have some kind of downtime in order to do maintenance and security updates but there are ways to coordinate that downtime so that it’s not visible to players. That’s one of the advantages of using a cloud infrastructure. Since we don’t own physical servers and thus are not limited by the amount of available hardware, if we need to do something like a security update, we can make new servers, install the update, remove the old ones from the matchmaking pool, and then destroy the old ones as they empty out.
Sounds like you have a good grasp on how other games have handled asset loading as well. One of the games that I worked on in the past actually had a system where they had overlapping areas between different game servers where the client would preload content from the server it was moving towards so that it would be able to seamlessly transition between them once it arrived at the transition boundary. Granted there were a LOT of bugs with that system that allowed for some hilarious screen shots. I wouldn’t be able to answer your question as to what we have in mind for how Star Citizen would handle these situations though, as that’s outside of the area that I’m in charge of. I am really interested to see how it ends up working though, because it will most likely impact how we manage our cloud resources as players move from one place to another.