The beauty of simplicity, an article in the November issue of Fast Company Magazine, addresses this problem, relating it to the success of Google, and changes Philips did to make usability one of the company's strengths on the market. Doing just enough, and no more, is driven as the recipe for success, and I'm willing to agree.
The Google front page minimalism keeps things appearing simple. It does not strain our limited supply of attention by introducing ads (a sure longevity mark of quality, by the way) nor any visible tools, except those used frequently enough to mandate being there. Additions do not enter lightly. While Google is a massive Swiss Army Knife under the hood, the front page does not expose every last tool to pick out the eyes of the majority of visitors who came by for something else. This is the "mostly closed" approach to the army knife, as compared to the "mostly open" approach offered by many competitors not keeping their front porch quite as tidy as Google's.
I find this a very compelling trade-off, but what makes me even happier is when there are hidden entrances everywhere to the tools underneath, for those of us mastering a larger subset of the knife than the parts visibly exposed. Without compromising with the needs and comfort of the many. I'm talking about things like query languages that allow me to look for pages across the web, where "ecmanaut" is part of the page URL, by prefixing
inurl:to the query. Or orthogonal approaches, such as moving in through different doors entirely, becoming part of the browser itself, as an even leaner search field at the top of my window, next to the browser menus. More power with a close to zero clutter footprint.
There is always a tradeoff between visual apparrency and hidden functionality, and in the above example I'm tapping into features of the Google advanced search page a mere mortal would rarely look for. The use case isn't large enough to warrant the clutter on the front page, or indeed even showing up on the advanced search page prior to a closer study and clicking-through of that page's select boxes, for that matter. This is good design, when you offer a wide feature set.
What is often even better, is offering a really small feature set, as does the iPod, for instance. It plays music. But in cases where the interface burden is zero or close enough to it, even the iPod goes beyond the trivial wide masses use case, and lets you store other random files on it when you plug it into your computer, as if it was your run of the mill memory stick.
It's important to get a good eye for when it is right and when it is wrong to add a new feature or whole feature set to a tool, product or application. We always walk a thin line between overladen user interfaces and not evolving to match the needs of niches and use cases we could address, when looking forward as visionaries to doing more than we do today.
There are many ways of evading the sharp edges that stick out when we add one more feature to the concoction; I'm quite fond of the self configuring Greasemonkey user script addition I recently did for my publish pinger tool, for instance. It adds buttons for sites you use to send notifications to feed readers and aggregators, but not any others. Silently growing added functionality in only the places where the user is mentally prepared for it, and would obviously benefit from it. And at no configuration cost, either.
This is exactly what makes Google move into the browser toolbar business with such decisive force -- the conventional web adds barriers between sites, and will only let you adapt to user net-wide behaviour through the small looking glass of seeing which page she came from to you, whereas a browser extension, or Greasemonkey user script, sees a larger picture, and thus is able to better leverage functionality fit for the specific needs of the user. Without spilling in to the browsing experience or comfort of any other user.
Living in concert with the browser is both a very large opportunity to do good things and an equally large responsibility not to do bad things. You need lots of trust from your user base to pull it off, and you have to earn that trust. Microsoft tried taking some first steps towards tapping into the possibilities that open up when you start looking at the web as being open to invasive composition, in introducing the concept of Smart Tags with the coming of Windows XP. This would have enabled them or third parties tying into the system to add content or functionality to any page on the web (such as adding context menus to certain links or words), much as the opt-in browser extension Greasemonkey does in Mozilla today. It received some really horrible press, and the public uproar made them abandon the project.
Google have so far not made any visible moves toward this, but has eagerly started to move into as many different fields as can possibly be fit under their corporate mission and vision statements (which is a lot); there is fortunately much on offer that could stand better organization and ease of access. They are wise to be cautious and equally wise to have and value their corporate motto of not being evil. Microsoft did not pull this through, I believe, mostly attributed to their lack of user base trust. I trust Microsoft to do what Microsoft believes they would gain wealth from, but I do not trust Microsoft to value my integrity above their own moneymaking opportunities.
Of course, Google is also a company interested in turning a profit, but they have done a much better job of maintaining at least the image of trustworthiness, when it comes to being trusted to value your interests. Not all feel the way I do about that, of course; there are many who will not touch Google Mail, for not wanting to place their pricacy in Google's hands. The same probably applies to your browsing, and running Google's toolbars, though there has not been nearly as much publicity on that topic yet. This would most likely change, in the event of Google's starting to visibly tap on the added powers of the browser extension.
The great strength in Greasemonkey is not only in the technology it offers, but also in the way that it does it -- each of these custom scripts requires the user to opt in, a procedure that is also very easy to do, once the extension is installed. You right click a user script, and pick "Install User Script", which is the first option of the menu for only this particular type of script. No annoying clutter footprint anywhere else, the gate to a whole different layer of web on top of the common web, and an immense opportunity to do new interesting applications that completely dwarf today's "web site mashups".
Greasemonkey is still in its cradle, as a phenomenon. It's very stable and rather mature development wise, and the user interface is getting better with every new release. You are still considered exceptionally front line if you have more than, say, thirty user scripts installed, though; we have about the same situation with User Scripts today as we did with browser bookmarks when the first Internet Explorer was designed, and the Favourites menu was deemed a suitable way of organizing them. (It works marvels for a handful, but loses usability steeply as you aproce dozens.)
Greasemonkey doesn't require you to invoke the scripts you install when visiting pages, though, so while its interface still has tricks to learn and sharp corners to shave down, the impact of scaling up the user interface is still quite low. I'm doing my share to help overcome these issues just by pointing at them, and the criticism is very well received. Development in the open when it works at its best.
On my own front, I'm eagerly pondering the possibilities of the Greasemonkey web-on-top-of-the-web, and have already done my own fair share of further Book Burro development to start the idea flow.
I hold a firm belief in the interconnectedness of all things. Greasemonkey, and other tools like it (Reify Turnabout adds the same kind of functionality to Internet Explorer, and the Opera browser has user scripts natively, though without the ease of installation Greasemonkey offers), empowers us to further this growth exponentially. Anybody anywhere can write the next big thing in site interoperability with no budget to start off from, and without the explicit help of the sites and parties involved. It can be shared with the rest of the world, just as easily and at the same lack of distribution cost -- it's just the matter of spreading the word. Publishing it at userscripts.org does much of that too, in a user scripts community that still has not grown very large.
Just never forget that to do and be any good, you must remember these two aspects: simplicity and trust. Toss either, and you will never take off, nor deserve to. Pay attention to both and strike a chord, finding just the right application, and you will most likely also strike gold. Today I do this for the sheer fun of it; I can not anticipate where this will lead to in a year, or two, or ten. It will probably be an interesting journey, though. And all who want to have front row tickets, too.