The trouble, of course, is that the tab key in the console is currently seriously crowded. Shift-tab toggles the console, tabbing while typing autocompletes, tabbing while typing an argument adds a space. Adding double-tab, which by necessity means i have to wait for a few milliseconds (more like 150 for my slow fingers to keep up) before committing actual tab actions reduces all the other tab functionality to a nagging crawl. What seems like a short moment soon becomes unbearable; in keyboard terms, 150ms is an aeon.

So how to implement it effectively? On one end, i’m actually really big on the idea of putting as much on the tab key as possible, making it the natural go-to-guy for your fingers whenever you want something done that ISN’T a command call, but on the other, i’m only up for that until it gets confusing.
Kieran/dBlue suggested an AJAXian approach where typing a word and then waiting a second or two automatically performs a search for you. I call this the “uhhhh-search”. I like how simple it is, but the fact of the matter is that the list of hits grows HUGE as soon as the search term is short enough. The idea is that searching for “cannon” will return both “bigCannon” and “cannonBall”, so a search of “ad” will typically give you a list of at least a couple lines. To have the console cluttered with random searches is not an ideal solution.

The final solution i’ve settled on is to drop any doubles or combinations, and try to be contextual. If the word under the caret is known in its entirety to the dictionary, skip to the end of the line and append a space as normal. If it is unknown however, do a search for it. I’m not sure it’s the optimal solution, but it leaves a minimum of clutter, can be toggled off, and still follows well with the tab-for-everything idea.

Give it a spin, and tell me what you think!

As a guy who has spent the better part of the last 2 years developing and focus testing “tactile” interfaces for software solutions (tablets and touch screens mostly, also some primitive mocap), i just have to throw in the problem that for a complex mechanism, in most cases intuitivity is actually counterintuitive. You’d think reaching out to touch a screen, “grabbing” an object and twisting it sideways was intuitive, but there’s an element of uncanny valley to it that actually throws people off. A lot. It also tends to alienate in-depth gamers who want functionality, not tactility. Latest example being Black & white 2, a game that has an interface designed around grabbing, touching, poking and throwing, and the more you play with it, the more you realise you’d rather just click on stuff than pretend to be this floating hand. Pushing the mouse cursor to the left on your screen to scroll left is, in software terms, THE way to go.

The whole grabbing and dragging thing is 3 operations instead of the simple 1, and what users want, in my experience, is to perform, not figure out how to perform. In simple game design terms, you have the zork school of text adventures, which would try to immerse the player by giving her a completely open range of movements, in theory. In practise it was a counterintuitive interface, kinda-fondly known as guess-the-verb.

Computer interfaces and human interfaces are awkwardly positioned. The lead criticism of Nintendo’s new motion-based revolution controller is that it places a lot of stress on the joints of the entire arm, not just the fingers, but on the wrist, elbow, shoulder, maybe even neck. Any intense game needs to focus as much interaction into as few movements as possible to be appealing in the long run, simply because, well, if you wanted a workout, you’d go to the gym. The wacko here of course is rythm based arcade games, ala DDR and Guitar freaks, but ironically, even in that area of body interfaces, the games that require the user to use say 6 pads instead of 4 and to use his hands to press buttons at the same time are the least popular. Simplicity is a virtue.

The tactility of an interface is useful inversely to how complicated the application is. The more you want a user to accomplish, the more you’ll have to deviate from truly intuitive movements like pointing, dragging, twisting and tapping, and get into esoteric things like drawing shapes/gestures and things of that order. Extending that functionality to body movement is actually a step backward. The ultimate interface is one that is easy to learn and that is applicable to any application after learning it. The keyboard/mouse is uncanny in its efficiency as well as its broad area of successful application.

An element of game design people, especially sim heads, tend to forget, is that games are entertainment to most people, and one of the few forms of truly interactive escapism out there;
If i want to be some crazy agile superspy, i’ll pop in Splinter cell 3, and step in the shoes of a character more able and more skilled than i am. The interface is a means to let me, the regular joe, pretend to be a veteran infiltrator and pull off the tough stuff with a measure of success. If the interface demanded a 1 to 1 mapping of my real skills into that of the game character, you’d lop off a good 99% of the market right then and there.

The abstraction in the user interface is what make games approachable; making the more direct or more “human” is, in my opinion, a strong, firm step backwards.

http://www.naturalpoint.com/trackir/

It looks completely idiotic. Looking one way to turn your virtual head, and thus limiting your field if view on the screen is the very essence of the human interface gone too far. When it’s directly detrimental to your gameplay, what’s the point.

To summarise, VR lacks success and support because it’s deeply counterintuitive and deeply weird to people with no interest in it, and because it demands custom interfaces for each and every application with no true time-tested design to it.

What i’d like to see is evolving the gamepad, to the point where controls are on 3 axes and we have actual depth control. I’d like to grab a force feedback bluetooth gamepad that gives motion feedback. If you press the button that represents a door opening action, and the door
cannot be opened, the controller physically prohibits you from pushing the button. Standing against a wall, trying to move against it can’t be done. The control stick won’t let you push that direction. Stuff like that, connect the player’s physical world to the virtual world of the
game. A good step was rumble, now couple that with temperature shifts (put a silent fan in there to cool it, already controllers that do this), buttons and controls that let you feel the texture and
composition of the objects you interact with. In an FPS, make the heavy weapons heavy. Make your fingers work to use them instead of imposing some virtual limit on you. The limit is real.

That, would be rad.

Suspension of disbelief is a tough matter, but closing the gap between the game world and how you interact with it is more about the human mind than some 1 to 1 proxy mapping, which felt weird in 98, and feels weird today. Abstraction and careful measurement of it is the bridge.