Shiny Chrome

Google announces Chrome a new browser they are developing. They even provided a comic book style explanation of the intended features, here. Chrome bulds on three big chunks of technology — Mozilla, WebKit, V8. Each is a particular advancement in their own rights.

Highly suggest one read through the comic book first to get a flavor. But here is what Google is trying to accomplish –

• Provide a multi process capable browser. The user can have several instances running under one user interface.
• The single instance of any process can be contained in a sandbox to jail any malicious code that the browser will detect.
• The browser will be inheriently ‘google gears’ enabled.
• The Javascript (aka V8) code can be contained within a process. It is not necessary for the script to complete before the browser is usable again. In some cases the javascript can be converted to what I presume is byte code for faster execution.
• The UI is rippable. Chrome will use tab metaphors. Similar to but no like how Firefox and IE do it today. Chrome intends to permit one to ‘rip off’ a tab from the UI and set it down on the desktop as a separate instance. Since each tab was its own process that is easily doable.
• A more intelligent autocomplete on the location bar.
• The browser will support a rich agent environment for the web.

Why?

To tell you the truth were I Google I would probably come to the same conclusion. The big three browsers — Firefox, IE, Safari — have essentially reached a impasse on futures. Its a browser after all for display of content and really no more. That Google was able to shoehorn the capabilities of Gears into to them is a testament to their ingenuity. But even Gears has its limits being a asynchronous transport layer.

I’ll give you an example. Some that from a human perspective looks simple — screen pops based on a incoming phone call. Practically every call center in the country relies on them. But my, go look at the technology used to accomplish it! A lot is curses based today. Of those that use a browser one must have an active instance open for the pop to occur, etc. If the browser crashes the entire session and data is lost. Worse you can only do one pop at a time, queuing the balance behind the IVR. Another words not optimal.

Google is proposing to do to the browser is like the difference between DOS and Linux. Permit the browser to work multi threaded and be able to react to differing types of content coming at it. They want a platform that is more malleable to a true client < --> server model. Possibly free of the restraints of REST and SOAP.

From a business perspective if Google can get inside of Microsoft’s IE development cycle it forces Microsoft to focus in an area not quite as profitable to them. Not only that but it tends to blunt the Software+Service model Microsoft has championed. If Google pulls it off they cut Microsoft off the net so to speak for at least 2 years. That would be the exptected development time for Microsoft to come out with a competing product.

Google’s biggest hurdle? Standards. Chrome is supposed to be inherently standards based. But Gears as far as I know is not part of the W3C standards nor being proposed for same. So it leaves a open question about ISO level adoption, be it Gears or Chrome.

Questions

Even with all of this. What about? —

Is XUL to be part of Chrome? If it is will Google assist in the development of XUL editors?

Is Google Gears to be extended or displaced in the Chrome development?

Is there to be a lightweight version suitable for Android?

Its Open Source. So will Google be teaming up with Mozilla to assure cross compatible coexistence where possible?

Interesting times ahead. We could very well see nettop boxes with no official OS but a boot routine and a Chrome based UI that handles everything else. Your applications lay in the SSD or the cloud even though they themselves are not web aware applications.