Maybe browser plugins are a good thing
I just read a blog post by Dave Winer. He is far too FF centric like so many people, but he has a point.
When you look at our ever more evolving web applications you can clearly see that this simply was never thought to be the task of HTML. Most of all apps are somewhat datacentric, and while with css you can quite nicely separate this content from its design it's still kind of botch. And at least when you look at apps like Google Spreadsheet you can see, that Javascript can do much, but it is too much constricted by the DOM and the whole way you have to do dynamic things.
So, obviously we have a problem there with currently only one solution: Plugins
But we don't like plugins. But why?
Well I think the main reasons are that they kill the open nature of the web. They are proprietary (what I personally don't find too bad), you don't know what they really do and you have to install extra software.
While reading Dave's article I thought, maybe the only real problem is the extra software thing. I mean, have a look at Silverlight or its open implementation Moonlight:
They built upon XAML, a XML derivate, and any programming language you like. C# is even an ISO standard. And then the whole thing is compiled and packed into ZIP file that is renamed to *.xap.
But now imagine the following scenario: We build the whole Moonlight package, including an compiler, into our browser, remove the compilation step from above, so that we still have the open source code, and just have to compile it on the fly, and then package everything up again in a ZIP file!
Think about it!
This only builds upon open technologies. We save bandwidth, because everything is compressed, it's all in a single file and we don't have to HTML format our data. And when you have a look at current Silverlight or Flash pages you can clearly see what a huge potential such a web would have!
Just think about it and tell me your ideas, maybe there is something I didn't think about...
experttease wrote:
You say there is a lot of promise in a web with webpages that Flash and Silverlight enable, but I make a point of avoiding many sites built with Flash because it is misused and clunky. Of course there are sites which need it like YouTube (only until HTML 5, hopefully), but on the whole these useful sites make up about 1% of the sites I've visited which use Flash (all in all I was probably happier with Media Player Plugin video, at the time when Flash wasn't prevalent elsewhere on the web).
I'm not looking forward to the encroachment of Silverlight either, as it is, as one person put it, the answer to a question no one asked.
All that said, I don't fully understand the technical implications of your proposal, and they may be a good alternative to the current scenario. But, I am more looking forward to the current approach taken with HTML 5.
One question: if such an agreement were achieved among browser vendors, how big a download/install would you expect the browser to be? (perhaps excluding the mammoth Safari 4). I have grown comfortably used to the small and light Opera browser.