This is not the typical sharding discussion thread.
Its about why they want to use it, and how that can be a problem.
Blizzard stated that, as they do not expect classic to have a rising population, they cant address the game as a usual new game release.
By that logic, it seems that they intend to increase the number of people one server can handle at once and shard it, so that after a few weeks/months after launch, the server will drop down to a more normal size and sharding becomes unnecessary while at the same time, they do not have dead servers they need to merge. Also, as many people in favor of sharding have argued, it would remove the need for que times as the servers can handle an seemingly infinite number of players due to sharding.
While that sounds quite logic and nice on paper, it also inherits possible problems that are worse than dead servers, merges and que times â besides all the orther issues that come with sharding that have been discussed over and over in the past.
-
If servers do not have a cap due to sharding, in order to not have log in que times, servers can get ridiculously huge, which would in turn make sharding mandatory. And if theyâd remove it still after a few weeks/months, the server would be overcrowded and in desperate need of adaptive re spawn rates. Either that or they do need hard server caps, which would invalidate the argument of pro sharders, that this removes the need for que times.
-
If the game becomes a huge success and the numbers do not dwindle as blizzard suspects them to, they will have to either split up servers in order to remove sharding, use adaptive re spawn rates and/or have to keep sharding all the way. With the first and last of those options being way worse than a few merges and/or dead servers (in my opinion).
My last point is a bit more guessing but I feel that its worth stating anyway.
In a recent Q&A with Ion he stated that they moved away from their previous server based design to the shard/phasing design of today, as servers are highly imbalanced faction wise (and i suspect population wise also) and they need it in order to make open pvp and other features of the game playable/fun for the faction that is outnumbered by a huge amount. This system in use has shown multiple times in the past that its NOT good at handling many players at one spot without sharding them, causing huge lag spikes and even crashes, going as far as to make GMâs plead to players to stop those kind of undertakings.
In direct comparison the cloud based server architecture seems to handle massive amounts of players at the same place and time poorly and/or even worse than the server based architecture they did use 14 years back. As a comparison: Server based architecture used by private servers today, which works with less than optimal scripting and emulations, does handle this specific case better.
Conclusion:
Beside all the other effects sharding does have on the experience and community of Classic as a whole (as discussed in many other threads), it also has (as stated above) other significant problems that make it a very risky system to use at all.
Id rather see them set for a specific server size with a server based system in place that uses modern hardware to increase the stability and performance of the game as a whole. With Blizzards polish in software and huge funds to buy/rent modern servers of that kind, big battles and/or events like the AQ gate opening shouldnât be a problem. While the launch will be less streamlined as to overcrowded starting zones and que times (and the need to develop some form of adaptive re spawn rates â at least for the starting zones), it would remove all the other risks that come from and with the sharding system.
Whereas one would have a worst case of long log in queâs, overcrowded starting zones, maybe some crashes in between and possible server merges, the other (sharding) option has worst case scenarios that would utterly destroy the whole idea of classic and with it⌠the game.
TL&DR: Scroll up.