Now that they all render so very well, what did you all expect would become the focus in the new browser wars?
They no they can no longer get away creating proprietary tags, DOMs and CSS properties (although I'd still like to see the W3C finally introduce <peek> and <poke> for HTML5), but what they can do from now on is mess around with the UI, and compete on that - and that includes coming up with proprietary features to cause lock-in for the user.
We're still in the phase where they're all stealing each others ideas (Well... mainly Opera's, including tabs on top, bookmark nicknames and searching from the address bar in version 4-5) but once all the 'standard' features are in all browsers, then you'll start to see radical changes to the browser user interface, including the removal of the back button. At least one browser maker will remove it, saying that Web 2.0 has made the concept redundant, and replace it with some new fangled mechanism, leading to its users becoming 'locked in' to their new philosophy. Then the others will all change other fundamental parts of what we see in browsers today.
Add to that all the peripheral services such as Opera Sync, Opera Unite, Widgets, Web slices, Firefox extensions, etc, and there will be much more lock-in than ever there was with Web 1.0 browsers. At least then all you had to do was switch browsers and worry about transferring your bookmarks, or change your website and pay the bill. In the future, users will be buying into a whole system, and it's not going to be so easy to change.
Makes you wonder why everyone (including MS) is so supportive of web standards these days. could it all be a diversion from what they're really up to? All these advocates thought web standards would bring us an open web, it may just be that it does the complete opposite - and browser makers are only standardising their rendering engines because it's all been done now, and there's nothing more to gain from competing with proprietary code. A bit like how cutting edge programming languages can sometimes be fractured as people compete on how to use the language until it all becomes rather passé. At which point, an ISO version is agreed upon by all and everyone sticks to that whilst getting on with the real business of competing with the programs they make using with that language.
We've passed the point where normal home PCs have adequate CPU and RAM. Making them faster and bigger is of no use to people anymore. However, it is providing a perfect opportunity to squander all those surplus CPU cycles by turning web browsers (dynamic document viewers) into pseudo operating systems (look how Opera now adds Widgets - through Add/Remove in the Control Panel!). The browser makers are happy because it makes them more important, their users end up locked in to a system and they make more money. And the OS vendors are happy because although they no longer get all the limelight, browsers aren't real operating systems yet, so you still need their product lurking silently underneath to make it all work. And the hardware makers love it all because it means they get to continue making faster chips and FSBs and bigger RAM sticks etc.
Seriously, what's the advantage to me of a Widget running on a widget runtime running on a browser running on an operating system? Why not save all those CPU cycles, not to mention watts, by standardising operating system APIs (a bit like POSIX but for all systems) and then having programs written in C++ or PHP or whatever language each developer finds convenient to them which can then be installed on a Mac, PC, Symbian or Linux, with the only 'system requirements' being the appropriate CPU architecture (Java has the same problem as the browser-OS model - it's too slow).