Monday, May 05, 2008

(Assignment #5) The shift from entity to tool

How will the web look in 5 years? There are two ways to approach this question: one is to look at what the newest technologies are, and then pick out which will live and which will die. The other approach is to identify a major trend. I will take the latter approach, partially because I just don't use the internet often enough to have a good idea of what the newest technologies are, and partially because I like abstract theory much more than practice.

First, let us recall how not the web, but computers, looked not five, but fifteen years ago. That old old time when people were using DOS and when GUIs were a new thing. Think of the way people interacted with a computer back then -- through a command line interface -- the user issues a command and the computer executes it. Interaction with a computer was like a conversation. But then came a fancy Windows GUI and everything was done by clicking on buttons. Now the computer isn't the other side of a conversation, but a place to put the tools you use in. Those tools being applications used to manipulate documents, files, and their subparts. So the CLI to GUI shift is from interaction with "the computer" to interaction with what it manipulates, data. I am not saying that one kind of interaction is better than another, but merely that what the user interacts with is different. But since the average user would rather directly manipulate objects rather than have a civilized conversation with an automaton, the GUI interface won in popularity.

Now look back at the shift from Web 1.0 to Web 2.0: In Web 1.0 we had static web pages which were reached by typing a URL in the browser. So we have the user request a page, and the browser retrieves it, and that's it (similar to the CLI interaction, isn't it?) In today's Web 2.0 we have shifted our focus, again, to the content on the web, as opposed to the medium through which it is delivered. What I mean by this is that YouTube is used to retrieve a video, Facebook is used to interact with a person, Wikipedia is used to get information, etc. This shift to viewing websites as tools becomes more apparent when APIs rather than the websites themselves are used, or when mashups are created.

But this shift is not yet complete, which becomes apparent during the interaction with these websites. When someone goes to Google or Ask or YouTube or Wikipedia the first thing they do is enter search terms, and then they get their data. Even when we talk about searching the internet we sometimes say "Let me ask Wikipedia." "Wikipedia," clearly, is an entity when we look at it like that, and the next step would be to make the search engine's presence less apparent. Today's browsers are already on their way to make searches more transparent, if you enter search terms into Firefox's URL bar, for example, it will sometimes guess and immediately call Google's "I'm feeling lucky", or sometimes it will send you to the search results. Plugins and programs like Mash Maker go another step and truly act as tools that manipulate the data on the web.

Another way in which the web is becoming a tool rather than an entity is through desktop applications that use the web without explicitly invoking the browser. Yes, these have been present throughout the history of the web, but with higher bandwidths and wireless internet in more and more places they are becoming much more usable and much more popular.

So in 2013 our interaction with the web will be a lot more transparent, and more intimately integrated with the desktop experience.

No comments: