Game Feature Ideas
SDxWiki

Many of our thoughts about the game come in the form of features we want to implement.

Continuing to assume a space MMOG with strong sim and RP attributes.

General Feature Ideas

Stuff not yet categorized...

Istvan Metagame information flow is a critical problem: there are no secrets on the Internet. Reconciliation probably has to be on the level of information duration: Keeping valuable information short-term escalates the information war in your favor over the fansites, which usually can't be updated fast enough to keep up with the changing world. Information that does not somehow grant advantage over other players would be important to make available freely, through in-game interfaces (further reducing the community recourse to fansites). DWM Aside from short-shelf-life data, abstracted data can also be kept in-game. For instance, you've gotten the cryptographic keys to a rival's comms. The game doesn't ever tell you what the key is, just that you have it. And of course, the game gives you tools for doing things with it, including (probably) sharing it with others. Istvan Perfect. Abstracted data, with in-game tools, fits this need very well.

Istvan If reactions were dependent upon a small number of variable player-object prestige or reputation attributes (in example, a low prestige with maintenance workers), the AI would only reference one or more player attributes to determine behavior. Keeping the number of these attributes small might be a necessary consideration. Detail is good for realism and immersion, but I've usually thought about this kind of thing on the government or organization level. DWM Yeah, this is a bit 'out there' in terms of programming complexity and, perhaps, storage needs. Not sure though. From a programming perspective, it's appealing to think of AI entities as independent programmatic units that have the same interface to the game that players do (at a binary level). This prevents AI 'cheating', keeps AI complexities from creeping into other parts of the game, and leverages off existing interfaces within the game program. (The latter point is fairly important; my preferred approach to organizing programming projects puts a big emphasis and a lot of effort into cooperative interface design. Individual programmers have a relatively free hand in implementing subsystems behind these interfaces.) Istvan I'd propose starting with AI on a "larger entity" - corp and govt management - level, but leaving "hooks" in objects such as stations and ships such that AI "operators" can be brought in as we are comfortable with the complexity. A simple early ship AI should be implemented for orbital shuttles, which are part "color", part economic visibility: since planets are intended to be outside the scope, shuttle objects would move station-to-surface and back and conduct transfer of commoditiies from stations. That transfer would be from the stations proper to a government's or corp's (controlling the plants) abstract inventories/repositories. Destroying shuttles would be both possible and have a direct impact on (A) resource movement, and (B) attitude of the owning entity toward the destroyer.