Finding things on the Internet is the most challenging and arduous task facing every person who brings up a Web browser. There are no road maps. We have no landmarks. We don't "live" in any particular place. The Web (and the Internet in general) is a foreign landscape that defies the sensibilities of the physical senses life and evolution have given to us.
You can have a start page but it's not really YOUR page. You have little control over where it is or what it looks like. You can sign up with a service and bookmark their "user-customizable start page" as your "home page". Or you can do like me and start with a "blank page". Write what you will in that browser space, it's yours to shape. But it remains shapeless.
Science fiction movies have shown us a vastly improved interface to the online world that we have yet to employ: a 3-dimensional perspective that treats everything as an object. We've seen it in "Johnny Mnemonic" and its clones; we saw a real-world example in "Jurassic Park" when the young girl navigated a UNIX visual file system. We see it in the most recent "Star Trek" movie when a Romulan miner throws a holographic image of Ambassador Spock's head at a Starfleet captain.
Google and a few other companies have proposed creating this 3-D interface for us using our cell phones of all things. Google's dream is to have you walk down the street and "see" (via a 3-D user interface) the businesses around you that have published information about themselves, their products/services, and other things online, on the Internet, there where the little computers-in-our-hands can grab it.
This 3-D interface would provide us with publisher-driven annotations -- but we would quickly want to make our own annotations. In fact, thanks to social media we annotate just about everything now. We Tweet what we're doing, we post where we are on Foursquare (well, some people do), we broadcast our locations so that our GPS apps can tell us we're in downtown New York City (as if we were not sure of that).
The future is here. We're just struggling with grasping it because it's so big. There is so much information about the world around us we haven't yet figured out how to organize and manage it. Traditional information retrieval, which evolved from crude metrics like Recall ("how many documents match up to my query") and Precision ("how many of these matched documents are rated highly relevant by some crude annotative metric?"), has failed miserably to provide us with the information we need NOW.
Bing, Google, Yahoo!, and several social media services have attempted to add a third axis to the IR space: Location. We now can ignore highly relevant documents if they are not in close proximity to our need -- assuming our need is based on where we are. Hence, it's easier to find a restaurant 3 blocks away just by asking our cell phone to find a restaurant than it was a few years ago.
But we're still missing a few metrics from our search tool repertoire. One very important metric might be called Chronocity -- the measurement of the distance in Time between where we are (Now) and where we were at some time in the past (or where we will be at some time in the future) with respect to a specific object, or document. For example, if we search for a restaurant within 3 blocks of our present location, and specify that it must serve Chinese food, we may find several appropriate restaurants. But will we find the specific Chinese restaurant where we celebrated our graduation from high school or college, our wedding anniversary, or a friend's birthday?
There are chronological factors that may alter our needs when we search and yet search technology (for that matter, annotative Web document technology) provides no way for us to add the chronological metrics into the algorithm. We cannot simply say to the cell phone, "Lead me back to where I was at 4PM yesterday when I bought a great hamburger from a street vendor". We have to make a note about where we bought the hamburger and then search for the note.
But that note won't be on the Internet. No matter how we think we play with the cloud now, we cannot (and in fact will not) store everything in our private lives on the cloud. But we may want to store these kinds of notes as incidental anecdotes in our traveling computers -- and subsequently have our search tools browse these kinds of content as well.
We could do that if we had a personal interface to our information that -- on the surface -- made no distinction between where the information is stored and when it was stored. Imagine yourself surrounded by a stack of semi-intelligent objects that have been labeled and cross-indexed according to what matters to you.
You tell the stack of objects, "I want to find the restaurant where I ate yesterday afternoon". All the non-restaurant objects immediately move away from you. The restaurant objects then confer among themselves and figure out which one was annotated with your visit. One quickly presents itself to you -- but if it's not what you really want you can sort through the remaining restaurants until you find the right one.
This kind of virtual search is impossible but only because we don't have the 3-D interface we need to both record/store information and manipulate it. Retrieving the information is the easy part. But using our position in Time to sort through it, that's still a function that is closed in our brains. It may take more searches on Google and Bing to find a restaurant you visited 3 weeks ago than it took to find that restaurant the first time around. The search engines know nothing of your personal experience with the restaurant, nor of your desire (perhaps unforeseen) to return to it.
So there is a chronological aspect to the Internet which is intrinsic and invisible. These are the "places" we have visited, the documents or widgets we have seen or used in the past. We try to find them again but we struggle to locate them. Their locations may have changed; their compositions may have changed; they may no longer exist. We have few tools for managing these things on "our side". We can bookmark pages in a browser but if you bookmark a thousand pages you create a search burden for yourself.
The Web That Was is often more interesting to us than the Web We See. That Past Web may still exist out there somewhere, but we are no longer in the same place relevant to that part of the Web we saw -- our personal Web space has moved, morphed, changed. We cannot take any part of the Web with us. At most the average person remembers a few dozen domain names easily. At best the average person relies on a small search idiom to find information over and over again.
When we set out to measure the size of the Internet we therefore cannot remain satisfied (for long) with a mere count of domains, hosts, pages, and components. We must also measure the parts of the Web that matter most to us. We will one day search for things we found in the past and fail to find them again, but they will still be there. They are simply farther removed from us in Time and there may be many distractions standing between us and those things.
If you could treat the world around you as if it were a semi-autonomous canvas and you painted really interesting or important objects in brilliant colors that would make it easy for you to find them again your personal search experience would improve vastly. But if you could name those things such that you call them and they come to you, your personal search experience would be a transformational function.
I want to go back to my favorite restaurant. I don't know where it is or what food it serves, but it pleased me more than any other restaurant I have ever visited. If I could only call to it across the years, and have it come to me, I would be happier still.
We must learn how to search through Time on the Internet for Time surely sorts and separates the information we find. We don't yet have the tools to do this, but we will.