Future Web

Today, when you visit a website you can click on pre-made links (usually handmade by the author) and you can scroll up and down. Up and down. That's about it.

Not much control.

When you play a modern computer game you can run around in a very realistic environment, shoot things, drive vehicles. Fly even. And talk live with the people you are busy blowing up.

That ain't fair.

Words are the symbols we use to think with right?

Even a paper book offers basic interaction possibilities unmatched by even the most powerful computer running a web browser. For example, you can dog ear a page (think of it as a local bookmark). You can underline. Highlight. Write notes in the margin. Call your friend up and discuss the second paragraph on page 36. You can't do that on the web.

The web, though simple to access and hugely useful in its ubiquity, has reached a plateau of interaction. Tim wanted you to be able to edit pages while browsing but Marc built an easy to use browser (is there something more to that name than meets the eye?...) which provides point & click interaction. Marc's browser took off.

Wind back twenty years.

Doug builds NLS. It takes a whopping half and hour or so to get to grips with. Instead of clicking on icons, one command at a time, Doug's system requires you to type (gasp!) commands in. Calm down, it was not like DOS. It was not like UNIX.

You would enter a full sentence, a bit like talking. You could enter something like 'open this document in a new window showing only the first line of every paragraph' by typing a few characters. If you should forget what you needed to enter you would not have to call tech support. You would move a finger to the question mark button, depress it and smile at a list of all possible, correct, commands. How could you fail?

Other advanced features. You could point to any object, any page, any paragraph, any word. Just like you can in a book. You could combine commands and save them (what some would call macros).

The movies make a big deal about speaking with computers. Speaking with computers is a big deal, but it's not the use of the voice which is the big deal, it's the opportunity to enter whole sentences and not just single commands like we do when we click on icons.

Picture using a perfect voice-recognition software package which understands you perfectly. But you cannot tell the computer to do anything other than the icons in front of you will allow you.

It would be the equivalent of travelling to a country you don't know well where the natives speak a language you don't know at all and you're armed with a translation dictionary which gives you the translation of individual words. You can only talk by saying one word at a time. Forget sentences. You could ask for 'coffee'. But 'please bring me a mocha to my hotel room in the morning at 7 but not on the weekend' would be way beyond your repertoire.

You may stay in this foreign land for the scenery, the food and the smiling people. But you wouldn't stay for the stimulating conversation.

Would you?

But that is exactly what you are doing when you surf the web. Sure, nice scenery, with Flash animation and QuickTime movies all over the place. Even CSS adding color to text.

Nice for when you want to visit Ironic Times for entertainment but not so nice for when you want to do real work and you have heavy texts to work your way through.

We have left paper behind and gone backwards. With paper you can, as Ted likes to point out; cut and paste. You can disassemble the work. Paste it together in new ways. Use different color paper for different sections or sources. You can lay it out on your dining room table and use the huge work space to easily move the pieces of paper around to get your head around it.

Of course, you cannot do that with text in a web browser. But enough of the moaning. Enough of the looking back to paper and to Doug's NLS.

What are the possibilities open to us now?

There are (at least) three possibilities:

1) Change all the worlds browsers.
2) Change all the worlds web pages to include advanced server-side interactivity.
Or...
3) Build a server based intermediary (Doug's term) which can serve web pages (first our own, then more and more pages) and add a little menu on key words which will allow the user to point to a word, select a command from the menu and have the command execute on the word (or phrase).

Liquid Information is about the last one.