What webservers are for browsers, are webservices for handhelds.
For the people that have been to TechEd in Barcelona this year must have seen the power of a rich client combined with webservices. (Filling in the questionaire after each session with the client app.)
This is where the real future of webservices lies. Why not build you applications with the power that client platform supports, instead of "Emulating" it on the white canvas called a browser?
This can only be possible if the .Net platform is being adopted like a browser nowadays is.
There is still one problem that the "magic canvas" has, and webservices have not. It's client control. When you are making a web application, you have complete control on how it's going to be displayed on the magic canvas. But when you have a webservice, you lose that most important control over your service.
Is there a solution for this? -I think there is and actually it's quite simple. "Put the magic back into the canvas." Let it imbedd your .Net clients like ActiveX controls do now. Why is this needed? People are used to the fact that a browser is their window to the internet. So they expect when you type the url www.myshop.bla you visit the online shop. How do you think people will react when they see a shop using all the power of their client?
Gereageerd door Montoya op :
You are right when you say we will go back to client-server computing But the Java community has tried this before. And it is not realy working now, is it? When you want this to work the CLR and .NET framework must be adopted like the JVM is, at least. And what about security, how many ActiveX controls do you decline! (I hope many for your sake).
The future is webservices definitly, but how do we get the "magic canvas" to enhance to this level? I don't know...
Gereageerd door NiMS op :
quote:Originally posted by Montoya: And what about security, how many ActiveX controls do you decline! (I hope many for your sake).
That's not really an issue here. You can ask yourself, how many programs do you download and execute? It's all based on trust. I was actually talking about the browser to make it easier for users to switch to running applications directly from their disk. But people are used to the fact that "browsers are the thing you use to connect to the Internet". This only has to be done, to make them switch easier.
The question I think you must ask nowadays, "Are the clients (pc's, handhelds) realy dumb?"
[ 15-07-2002, 16:38: Message edited by: NiMS ]
Gereageerd door Montoya op :
The clients aren't dumb these days, I would say.
But that does not get any kind of platform onto all devices does it now?
I've asked myself five years ago: Why do we accept this stupid browser as our UI on the internet?
The simple answer then was: Limited speed and reliablity of connections.
But why did it take over five years to realise we are back at square one. Networked computing only the network now is the internet.
A simple answer to this question would be: We were affraid, very affraid! Linking up to such a huge network for one. And second, but the most important, the big guys did not want to coop in make standards!
Now here we go, and things are getting there, aren't they?
The big issue here is will the big guys coop on making a good rich client platform! Of course based on webservices. But I don't think we are just gonna download our entire enviroment from the web.
![[Smile]](smile.gif)
But the Java community has tried this before. And it is not realy working now, is it? When you want this to work the CLR and .NET framework must be adopted like the JVM is, at least. And what about security, how many ActiveX controls do you decline! (I hope many for your sake).