Time To Take the Internet Seriously

santosh maharshi passes along an article on Edge by David Gelernter, the man who (according to the introduction) predicted the Web and first described cloud computing; he's also a Unabomber survivor. Gelernter makes 35 predictions and assertions, some brilliant, some dubious. "6. We know that the Internet creates 'information overload,' a problem with two parts: increasing number of information sources and increasing information flow per source. The first part is harder: it's more difficult to understand five people speaking simultaneously than one person talking fast — especially if you can tell the one person to stop temporarily, or go back and repeat. Integrating multiple information sources is crucial to solving information overload. Blogs and other anthology-sites integrate information from many sources. But we won't be able to solve the overload problem until each Internet user can choose for himself what sources to integrate, and can add to this mix the most important source of all: his own personal information — his email and other messages, reminders and documents of all sorts. To accomplish this, we merely need to turn the whole Cybersphere on its side, so that time instead of space is the main axis. ... 14. The structure called a cyberstream or lifestream is better suited to the Internet than a conventional website because it shows information-in-motion, a rushing flow of fresh information instead of a stagnant pool."

Comments

Re:condition: buzzword alert

[zappepcs]

Exactly, kind of, I think.

The Internet is not a thing like the 'winter olympics' or recording industry. The Internet is the system of communications systems which allow the transfer of information (as well as aggregation, falsification, and overload of). It changes the source of information for those who regularly access it when compared to the time before the Internet.

What needs to be discussed is not cyber this, or virtual that, but how users use information. Lets face it, for a large portion of the population the phrase 'use information' is rather optimistic. Aggregating information, presenting it in a way that is both intuitive and useful is something of a holy grail. We've seen many attempts to do things like this, and each of them has their fans and foes. What is being suggested is essentially that we all need to have one set of cultural values. Looks good on paper, but it makes a huge mess of things in real life.

Then again, look at Microsoft Windows. How may people do you know that think this is how computers are supposed to work, and anything not like Windows is weird?

A single cultural viewpoint is wrong.

So when did this guy predict the internet?

[rossdee]

So when did he predict 'the internet' ? Was this before or after Al Gore invented it?

AFAIK Shoghi Effendi predicted the internet back in 1936:

"A mechanism of world inter-communication will be devised, embracing the whole planet, freed from national hindrances and restrictions, and functioning with marvellous swiftness and perfect regularity."

Re:Time-Based Filesystem

[PaintyThePirate]

Interestingly, this is the approach that OLPC and now Sugar Labs have taken for file access in Sugar, using the Journal activity. This is also the direction Gnome is heading in, with Zeitgeist and its GUIs.

It's a little strange at first, and it certainly can't replace normal file browsers completely, but it ends up being pretty convenient in day to day use. Of course, these aren't filesystems, just layers atop them.

Re:Dear software engineers

[AuMatar]

I'd love to know what you propose would do better and still scale to tens of thousands of page requests per second, and can deal with malicious network nodes and nodes dropping off the network without notice. You do realize that TCP is also doing sessions on top of a sessionless protocol, right? Is TCP poorly designed?

Doing sessions on top of sessionless on top of sessioned is poorly designed. That's the current situation- interactive apps written over HTTP on top of TCP. HTTP is a good file transfer protocol, but it doesn't fit the modern usage of many webpages, its being shoe horned in because everyone uses browsers and that's the only way they communicate. It's past time for a new protocol at the HTTP layer made for web applications that can co-exist alongside it.

If the web was designed to be pixel-perfect, browsers would be as messy as Win32, trying to maintain backward compatibility with all sorts of different displays. Either that, or everything would be monochrome at low resolution.

But that's what every damn web designer wants, and what they struggle with HTML, CSS, and Flash to achieve. From frame hell to the equivalent in CSS, they design it assuming that it should be pixel perfect. It's time to educate them or give them what they want, the current hacks they use to try and make it so are a huge waste of time and money.

The only language I can think of that has arbitrary functions like RPC built-in is PHP. If you think PHP is the epitome of language design, then we have nothing more to discuss. Most good languages separate the language itself from the standard library.

You don't know many languages then. A good language for the web would recognize that it's client-server, and provide for built in ability for automated data transfer and calling of functions on the server. Instead we have the steaming pile which is Javascript, a bad language to begin with, married to the utter hack that is AJAX.

Sure, on the whole, it's not the best that we could do, but if you think nothing about it is well-designed, well, what would you propose? Flash?

The web wasn't designed for applications. Start over. A new transfer protocol based on sessions. A new display format based on SVG or similar technology with access to all common widget types (menus, sliders, combo boxes, list boxes, other things that the current web can't do well or at all). Scrap js and use a well designed language, one that's tier aware. And make browsers able to use this format or the original http/html stuff, as was always intended- that's why URLs start with htttp://.

It'd be a year or two to work it all out, during which we'd continue with what we have now. The end result would be a huge increase in productivity and ease of use, since we wouldn't have to wedge around broken protocols and throw in hideous hacks.

The last thing anyone needs is Microsoft reinventing the web.

As developers of the most used web browser do you really think they could be left out of anything? They shouldn't control the process, but they need to have a seat at the table.