The blues have said the problem is that they’ve hit the limit of what hardware and platforms can do. Based on that, it’s becoming more and more obvious that most of the human technical resources allocated to Classic are on the design side than on the engineering side. Modern WoW was engineered to horizontally scale using cross-realm and sharding. Retail sacrificed realm-identity in order to follow that path. Now we have a conflict between Classic’s design philosophy and Retail’s architecture. The engineering path forward with which they have become comfortable is fundamentally incompatible with the user experience they are trying to replicate.
So they throw their hands in the air and begin broadcasting two absurd claims: (1) they’ve hit the limit of what technology can do; and (2) since there’s nothing else they can do, only the players can resolve it, so the finger must be pointed.
The latter point cavalierly disregards a valid conclusion on the topic which has been reachable for a while: monetizing a bad solution has not motivated them to be good stewards of realm population. Every time they fail to act in accordance with the outcomes they’ve had more than enough time to acquire data to anticipate, they are rewarded with all the revenue from paid transfers purchased by players struggling to correct the population issues.
The former point is what leads me to the conclusion that we’re dealing with designers instead of engineers. Hitting the limits on sockets per network interface? Bottlenecking database connections? All of these issues they’re raising are indicative that they see realms as the basic unit of horizontal scaling (like it is in Retail) and that it cannot be done in any other way. But it most assuredly can… by engineers. No matter how intertwined and spaghetti-coded the realm services might be, they could always be subdivided as necessary and employ IPC internal to the hosting environment to scale a realm to support more and more connections.
But that’s hard. And expensive. And contrary to how Retail is architected. People would need to be trained on new methods of deployment, configuration, and maintenance. New points of failure would be introduced. It’s legitimately scary. So they’re going to resist it. They have reasons to do so. Unfortunately, it happens to be the case that: (a) they are responsible for allowing things to get this bad; and (b) they really could do something about this on a technical level, it’s just difficult and pricey.
They’re in a complex quagmire. It will be interesting to see how this plan of blaming the players and adopting a wait and see attitude pans out for them.
Signed,
An “arm chair” Enterprise Software Architect
P.S. I see you, Classic Team. For what it’s worth, you have my sympathy.
the guy who blue posted just did it to make himself feel good. Calling customers “arm chair psychiatrist engineers” and saying “just transfer bro there is no consequence.” They have the technology look at retail lol. Cross realm solves this in a day. The fact is they either don’t want to spend money on classic or the team is just incompetent. Probably the latter .
Well…as someone who is actually a Lead Integration Architect, Lead Systems Engineer, and Azure Cloud Architect…I can fully appreciate AND understand what Aggrend was saying regarding technology limits.
SO…threads like these, created by people who really have proven to have zero clue, are exhausting.
Don’t forget that they already destroyed realm identity with all of the transfers, paid or otherwise. Exactly zero realms have the same “identity” and community as they did in early Classic.
Unfortunately, we’re not going to be able to resolve our particular dispute in point of fact without access to the code for the services. But, in the abstract, of course there are limits on single channels all over the place, but the way to get around them is by taking a monolith and splitting it into pieces. If your argument is that it’s not feasible for them to do so, fair enough. I came to the conclusion that it might be so, myself. But if your argument is that it’s not possible to do so, you’re enabling them to hide behind technical constraints to avoid the accountability for letting things get this bad through mismanagement of the population… a point your reply conveniently did not address.
I’m actually confused if you are stating that this is the limit of technology or just their technology. It may not be possible on the hardware they are currently using, with the current systems in place, but saying it’s impossible outright is just not true. Nothing OP said struck me as exhausting like the other posts where people just say “make server bigger”.
I posted this elsewhere but it didn’t get any traction, nonetheless I think it’s relevant to the issue OP brought up so I’ll leave it below (it’s a round-about solution, but it IS a solution - and, further, it can be implemented with existing hardware).
Credit to user: glutchpls on reddit
Possible solution for queues: Temp realms
For all mega realms or realms with queue, spin up a completely new server named 2, (e.g Benediction2) where names and guildnames are shared and taken, and free transfers are always available between them, with the promise of a merge once queues are gone after wotlk launch.
This completely removes the fear of a dead realm
People would transfer immediately to the new realm to level up, especially small friend groups, solo players and new players, reducing or removing queue
“Shared/taken” names and guildnames between the 2 servers completely removes merge conflicts and transfers has no issues
Free transfers between server 1 & 2 at any point removes the fear of splitting up guilds or friendgroups
Removes the need to lock character creations which blizzard just did on mega servers
Cons:
Available names could be harder to find, but probably easier than on retail which have had names reserved for over 15 years
What if server2 gets full + queues and they can never merge? Spin up server3. But if this keeps going, blizzard might as well enable free transfers between all servers (not between fresh and non fresh). This would also solve queues, but remove a small income from blizzard, which they might not like. But at the same time, they might earn more from people not quitting due to server queues
Perhaps limit the free transfers to once every week to prevent too much AH shenanigans
I’m thinking this would be reasonably easy to implement with the only downside being there would be a little less money from $25 transfers going into blizzards pockets - upside of course being less (or zero) people in queues and more playing the game they pay $15 a month
