Feedback:User/Jette/Extensions & plugins

Extensions or plugins or userscripts or whatever you personally like to call them are popular. Firefox has proven that, as have a plethora of other applications. They make a program more comfortable, visually appealing or easier to use for the person using them, without forcing personal preferences down anyone's throat. Obviously, plugins that affect the connections between the client and the server would be difficult to manage and allow for a lot of abuses, including more efficient bots & other issues. User-made plugins that affect the interface, commands and other similar aspects of the client could be quite useful, however; a few examples might be a plugin to replace the default chat interface with something more akin to a proper IRC client, such as irssi or even a more powerful text editor like vi, or a minimap plugin that replaces the default textures with real-time high-altitude top-down renders of the battlefield (for those with GPUs that can handle it). This would virtually eliminate the need for programs such as TexMod as well, but ANet could claim that they cannot assist support issues related to plugins to reduce the influx of people who do not understand what they are doing breaking their client (and, of course, the simplest solution and quickest response from support or forums is "unplug the plugin" to fix problems).

There might be several disadvantages to this system: the first, for example, is that it could be possible for players to create malicious plugins that log passwords and send them to an outside source. This could be solved quite easily by preventing plugins from loading before a character is selected, making it impossible for them to run while the e-mail address & password are being entered. Another potential issue is the ease with which such plugins would make it to use bot programs or other mass-farming tools, but this point is largely moot, as all the tools necessary for an efficient bot are already available in GW1 through the use of an outside program. Therefore, using a bot via plugins would be no more efficient than using a simple mouse/keyboard script, and if anything, easier to detect. Other than modifying client-side elements such as renders, terrain, lighting, interface, etc.; plugins would be incapable of performing any task a player themself could not perform.

A final disadvantage is that the program would be slightly more "open" to players, i.e., in order to know how to force the engine to load an alternative model or txture, the player must have at least rudamentary knowledge of how the model or texture system works. Although it would hardly be sufficient to support reverse-engineering or compiling an original copy, and of course no server-side information need be exposed, it is still a point of concern. This could be seen as a positive or negative aspect, depending on your point of view.

Although at first it might seem unlikely many players would be interested in modifying their client in any way, the immense popularity of plugins, widgets and other similar mini-programs for other software indicates that this is not so. Even the TexMods available for players now are popular amongst players, both amongst "ricers" who add pretty interfaces to their client, to the more utilitarian purposes like more readily identifying a skill animation or helping with map exploration.

Personally, I would recommend Lua to extend the client, as it is a simple scripting language designed with application extensibility in mind, and has a proven history of effective use in video games. A more immediately familiar language such as JavaScript could be used instead, if a wider range of potential developers is desired.