I really haven't looked deeper into it. I'm basing my statement on the notion that HTTP was invented way before the internet went mainstream and that at today's pace of technological evolution every passing year takes us exponentially further away from the original purpose of HTTP.
I would be hard pressed to believe that Tim Berners-Lee envisioned HTTP becoming as popular as it is today and thus designing for this popularity in anticipation. I would be more easily convinced that HTTP is optimal for the time and purpose for which it was designed. Like starlight, we are still within that scope of possibilities, although some VERY creative solutions are available, such as Ajax - but this just indicates to me that needs of the masses are way beyond the needs for which HTTP was originally designed.
Using this approach, it would seem perfectly logical (in my mind) that the needs of today's web be looked upon with a fresh attitude as far as what we use it for today and what we may use it for in the next 10 years. Perhaps we need applications that are capable of dynamically acquiring new protocols (this would be a nice paradigm shift enabling so many possibilities) and thus enable a new level of interaction with the web than that to which we are currently constrained. By unhinging the browser from HTTP then there is an incentive for people to develop their own protocols thus opening up the full range of possibilities for people to invent the future web, whether it be semantic or something equally fantastic.
I will state out-right that I am merely dreaming about possibilities here and that I have not really given this much thought - there are easily many more people that could approach this problem and come up with better solutions than I could; I'm just saying that the future web deserves something better than HTTP. I would also propose that perhaps the web needs to be viewed as a multi-web when considering it's role in the future.
I would be hard pressed to believe that Tim Berners-Lee envisioned HTTP becoming as popular as it is today and thus designing for this popularity in anticipation. I would be more easily convinced that HTTP is optimal for the time and purpose for which it was designed. Like starlight, we are still within that scope of possibilities, although some VERY creative solutions are available, such as Ajax - but this just indicates to me that needs of the masses are way beyond the needs for which HTTP was originally designed.
Using this approach, it would seem perfectly logical (in my mind) that the needs of today's web be looked upon with a fresh attitude as far as what we use it for today and what we may use it for in the next 10 years. Perhaps we need applications that are capable of dynamically acquiring new protocols (this would be a nice paradigm shift enabling so many possibilities) and thus enable a new level of interaction with the web than that to which we are currently constrained. By unhinging the browser from HTTP then there is an incentive for people to develop their own protocols thus opening up the full range of possibilities for people to invent the future web, whether it be semantic or something equally fantastic.
I will state out-right that I am merely dreaming about possibilities here and that I have not really given this much thought - there are easily many more people that could approach this problem and come up with better solutions than I could; I'm just saying that the future web deserves something better than HTTP. I would also propose that perhaps the web needs to be viewed as a multi-web when considering it's role in the future.