Two Web Fixes

here are plenty of things wrong with the web. (We can skip the right things because nobody gripes about 'em. ) Here are two solutions to free up bandwidth and modems ... easy fixes. The basic tactic in each is "pay in proportion to use".

1. Fixed Rate ISP's

AOL made a huge mistake going from 20 hours for 20 bucks to $20-unlimited access. No one has any reason at all to get off the line anymore ... hence their modem problem. This is also unfair since AOL must turn a profit (it's a business remember?), twenty dollars to everybody means that those who use the internet least are underwriting "chatters" who are a NET LOSS to AOL. Result - AOL survives on the backs of short time users. They changed their rates under pressure from other ISPs who offered the same deal.

What they should have done ...

is what every ISP should do. Namely, charge about 50 cents per hour of connect time ... nothing else (except for exclusive optional content).

So you're on the net for 49 hours this month ... hmmmm ... $24.50.
5 hours = $2.50
328 hours chatting = $164.00

See what I mean ?

Mr. Chatmandu pays for his own use and has a minimal incentive to get off the net when he takes a shower. His use is not subsidized by the other guy who just wants to pick up his email. And 50 cents is not fixed in stone. The rate must be adjusted so that the ISP meets his desired profit margin. Whatever that rate is, it will be ... fair to all ... responsive to standard market forces ... and ... will give everyone some reason to get off and let somebody else on.

2. A Real Search Engine

What? ... What is Alta-Vista? ... What is Excite? ... What is Yahoo? ... What is Lycos? ... yada ... yada ... yada ...

I will tell you what they are in one word ... garbage.

And in a few minutes you will see why they are garbage and belong in the same pail with the collected works of Picasso.

LIBRARY ... card catalog ... 100s of millions of man hours of development ... gone.
REPLACEMENT ... computer terminal ... "tunnel vision".

The greatest loss in the history of data storage & retrieval ... not a whimper of protest (in public). The word "serendipity", I think, may have been made up to describe one's experience in the library ... in card catalog days. When I USED TO go to the library, I USED TO have something like this happen EVERY TIME. "Oh, look there. The book I was looking for last week ... what a 'coincidence' to find it here." No coincidence ... human management of a human resource (at its best).

No more. Now, I put up on the screen just what's expected. And it shows whether the book is checked out or not. Isn't that wonderful? Now I don't have to worry about finding something I wasn't looking for. Thank you, Mr. Computer Programmer, for saving me from that awful fate. Now if only someone could invent a pill to keep my mind from wandering and from daydreaming ... I could just get down to business 100% of the time and be 'more productive' ( ! ) Someone should excavate Commander Dewey and check for signs of rotation.

What's so good about a cumbersome card catalog?

It was cross indexed by millions of real people constantly tinkering, reinventing the thing, expanding it, putting the totality of their subconscious experience into the "box". Tunnel vision is the name I give to its replacement ... a system designed by two geeks in a garage in Hoboken last Tuesday night.

Newsflash !

The rigid software of a computer cannot possibly compete with the "compliant" software of a human mind. We resonate. Machine-human symbiosis will take much more to achieve than has yet been done. To make a machine anticipate the desires of a human, that machine must become a defacto human. It's not gonna' happen ... not in our lifetime.


A valid search engine (valid for human interaction) requires human constructors at every stage. No machine help will facilitate a "human" search.

But the Web has hundreds of millions of documents. How can one hope to deal with them without machine help?

Two words ... Paid Indexing
Note: Pay to index ... Free to search.

To hand index millions of documents requires millions of human indexers. And ... they must be interested in the project. Where do you find these millions? Obviously, the people who make the pages are the only ones who can do the job by hand with any quality control.

A search engine must be a two dimensional map of web sites on several levels wherein a person who wants his page found rents for a nominal charge ($1 per year?) a pixel in which to store ... a name | a color | a URL | a synopsis (2K?). If you want to cross index your page ... buy another pixel and put it elsewhere. A 'bot fetches your page one time to check the URL. After that ... nada. No one gets erased from the database because a server is down. You paid for a year ... you get a year.

Wheat is separated from chaff. Money talks, BS walks. And how about a business rate on a business search map? $10 per year per pixel ... that'll keep out those "pesky personal pages".

Imagine the opening page of such a search engine. Only big corps or government institutions will be visible (have enough colored pixels to show their name-logo on the top level). Hit the space bar and toggle to a lower level ... space bar again and yet another level ... maybe three or four levels max.

Click once anywhere on a map and that place moves to center screen (this is a method of moving around on the same level). Click on the centered selection to go to that page (or general map area). Page name and synopsis is readable on lowest level.

This is enough hypothetical design. I think it is clear to you by now what I meant when I said that present search engines are 'garbage'.

One last design element: Go to "night mode" and type some meta-word description and all the related files show as bright dots. Find (visually) where your kind of pages hang out. And, since this engine only indexes ~2K description of a page, it's very, very fast. And if you find what you want fast, aren't you saving bandwidth too?

EBTX ... "man on a million missions"

Ebtx Home Page