On 2 Jul 2010, at 11:39, Patrick Durusau wrote:
>
> Good point. But the basic tools to handle data have been around for a long
> time.
The web could only get going in the 90ies when
1) Windows 95 become (A GUI) widely deployed and relatively stable and had
support for threads
2) modems were cheap and available
[3 the soviet unions had fallen, so the fear mongers had no security buttons
to press]
In 1997 the SSL layer (https) gave an extra boost as it made commerce possible.
> Why so long to get to the place where users can say: "I want to make one of
> those." ?
There are many reasons, but most of all is that people don't in fact
understand hypertext, as being linked information. Or the people in charge of
data don't think of it that way easily.
Engineers have for 50 years been educated in closed world systems, every
programming language
including Prolog and lisp have local naming conventions that don't scale
globally, and database people
make a fortune with SQL. The people interact only very lightly with the web.
Usually there is a layer of "Web Monkeys" in between them and the web.
So when you ask those engineers to build a global distributed information
system, they
come up with the closest to what they know - which is remote method calls - and
they invent XML/RPC which leads to SOAP.
So it is not easy to get the knowledgeable people on board. The Web Monkeys
are not very good at modelling, and the back end engineers don't understand the
web. Finally the business people have problems understanding abstract concepts
such as network effect.
It just took time then to do a few demos, which the University of Berlin put
together, slowly getting other people on board.
It just takes time to rewire the brain of millions of people.
Henry
