Ask HN: Why does it feel like MMO server tech hasn't changed?

5 points by Daegalus ↗ HN
So this comes from the recent Release of New World from Amazon. They have a hard cap of 2k players and 2-3x that per server I'm queues. They are adding more servers often but usually not enough.

Now, I love the MMO genre and have been through many MMO launches, and this isn't new to me. But as a DevOps engineer, I wonder how we haven't improved server tech to minimize or remove this?

I don't know much about MMO servers out outside of the one I'm building for an old MMO, but I feel like a new modern MMO should have some innovations to solve server capacities or so many servers.

Why does it feel we have made no progress towards limitless servers, or single world games where you have fleets of servers handling 1 world? Why are we still separating players by servers and not employing phasing to help with clutter of people seen.

Or move characters to be disconnected from the world they play on. So I can freely choose the server I play on regardless of my character. Server choice can be more easily changed without starting over.

I understand the complexities of latency, server to server syncing, database capacity limits, etc. But I feel like those actually making MMO servers for a living have the expertise to make this better.

And then there is Amazon, they have the entire power of AWS to add servers quickly. I don't get how they can't have scaled extra servers on day 1 and day 2 with some cloudformation, terraform, docker, kubernetes, or whatever is the right tool for quick provisioning. They are adding 20 servers at a time. There are more people queued than playing atm.

Just want to start a discussion to see more view points, maybe some expertise of those working on MMO game servers, etc. There is obviously things I'm not aware of or not taking into account, but I would like to know.

6 comments

[ 2.9 ms ] story [ 33.0 ms ] thread
Money.

I'm sure all the issues you cite can be solved, but it costs money to address them. How much are you willing to pay to play an MMO, either up front, subscription, or free-to-play freemium?

Thats fair, I mean I don't mind paying a subscription. I can understand the money aspect. I just feel Amazon has the resources to throw money at the problem/development to begin with compared to most developers.
> Amazon has the resources to throw money at the problem

Amazon is a company motivated by profit. No one at the company is going to throw money at a problem unless solving it will be profitable. New World standard edition is $39.99, with no subscription fee. Blizzard used to charge a small fee to buy the game, but the monthly sub cost was what sustained it. They recently dropped the up-front cost. Users pay per month and play, and there are plenty of additional premium things to buy, fortunately they are all aesthetic, compared to many free-to-play that end up being pay-to-win.

It's unclear how Amazon will make money on New World, although I strongly suspect it will involve collecting and selling data about the users, along with some kind of microtransactions. I'd guess freemiums will come along later, too.

In short, just having a lot of money is not sufficient. For a business to want to solve the MMO scaling problems, it has to find a way to get some return on the investment. MMOs just aren't All That.

Very fair assessment. New World had a cosmetics shop already with equipment skins. so they will probably monetize that way too.

Thank you for taking the time to write that out.

This is their first MMO. They're playing it safe from the server tech side but I agree they seem to have the same limits as the very first gen MMOs like Everquest or Asheron's Call. Player names are global unique so there's probably some plans to unify servers at some point. World size probably also played a part as it is on the smaller side.
I do think the world set idea is smart. It will make server merges safer and easier later, its probably why they dont mind adding 100s of servers now.