Re: gforth webserver, why isn't forth used all over ecommerce?

helmwo@xxxxxxxxx wrote:
Forth is really too simple. Most people expect something like Java
or .NET-framework. They do not see that they write code with an expiry
date. Since Forth is so simple there is no expiry date - if a word is
missing, it's not complicated to implement one usually. It's also not
complicated to write systems that support multiple levels of
compatiblity. But in a hurry business like web-apps nobody cares about
(it even will ensure jobs in future).

That's not it either. Nothing prevents people from building up their own sets of abstractions, utilities, and so on on top of environments like Java or .NET. Just because Java provides a collection class doesn't force one to use it-- and in some cases, programmers won't use it because they have specialized requirements. Just because .NET supplies high-level serialization functions doesn't mean you're forced to use them-- you're free to write your own if you have special needs.

No, the real issue with why Forth isn't used as much is something I touched on before. People want to spend their time solving interesting problems, and endlessly reimplementing jellybean code and common utilities isn't interesting.

Which would you rather do-- reimplement something that dozens or even hundreds of programmers have implemented before, or work on the "meat" of the problem? Which would you find more satisfaction in-- implementing yet another minimal HTTP server, or designing the application that rides on top of that server?

There is a time and place for everything. There are cases where reimplementing a wheel makes sense. But the problem the Forth community seems to have is they don't know when it stops making sense. That's seen in the antagonism against libraries and the elitist (in the bad sense of the word) attitudes against programmers who choose to specialize or who have different core competencies they draw from.
.