oslsachem - 2012-02-06

These are my observations about the suggestion of keeping a list of active stendhal servers to help to build a community around the project.

Why keep a list of active servers?
** To attract potential contributors who play in a server hosting a (possibly heavily customized) version of the game: the servers hosting the game should refer players back to the project. But there is no way to enforce this inside the game (i.e. in a way highly visible to players) because they are free to modify the client and server files and strip out any reference.

** To get feedback that helps to improve the game:
- client-side feedback isn't guaranteed in customized servers as stated above.
- server-side feedback, on the other hand, comes almost exclusively from users who host unofficial servers.

** To shift, in the long term, the responsibility of keeping a gaming community to user-managed servers so that developers can:
- avoid the time-consuming task of supporting players and its potential burnout.
- focus their time on the management and development of the project.

The benefits of having a list of active servers as a convenience for players:

- Offering a server which is close enough so that latency doesn't render the game unplayable ( there is a list of possibly unplayable regions from the official server: http://just-ping.com/index.php?vh=stendhalgame.org&c=&s=ping! ).

- Offering a localized/customized version of the server which uses a player's native cultural references or the players of which use mainly her native language or share her time zone.

The drawback of using a list of active servers:

- The project ends up unofficially endorsing a group of servers operated by people who can neglect their players ( people don't like investing time in an online game which isn't well moderated). Therefore, this list would need at least a reliable ranking system driven by player votes.

- A player may suffer server lock-in if she decides to switch to another server but wants to keep her character.

This on the other hand puts into question the need for keeping an official server to:
- guarantee user feedback and attract contributors to improve the project.
- keep a demo version of the server for the users to try with their clients.

Can user-managed servers at some point assume these tasks?

Can user-managed servers be sustainable and trusted by players?

Does keeping an official server make the project compete with a part of its users (namely the users that host servers)?

Is the project interested in the proliferation of user-managed servers or just in the distribution of the clients?

Should the project assume any responsibility for the security risks of a user setting up a server or can the project be discharged from that liability?

Should the setup of the website part of the game server be better supported?

In this regard, would it be feasible to use a java-based web server together with the h2 database to replace Apache/PHP/MySQL for zero-configuration and easy deployment?

What would the need for load-balancing (a technique presumably used to avoid userbase fragmentation) be in an already fragmented userbase scenery of a list of servers instead of a single official server?