The Connected Circuit

So earlier today Google decided to announce a bombshell. That they have been in development of an open source browser to compete with the very competent likes of Mozilla Firefox, Microsoft Internet Explorer and Opera. So why you say would they ever do that? Aren’t they purely a web-centric software company? Well to answer the second question, they were. Google has been the big name in search since the earlier 2000s, after the big tech bubble of the 90s, but they are now providing a wide array of applications, from apps on phones (Android) to ones on the desktop space (Google Earth, Google Desktop) and your traditional things, such as email (Gmail) and mapping (Google Maps). But to handle all of this, they decided that they needed to create something that came cope with the increasingly important dynamic webspace that is the interwebs, so what was created was Google Chrome.

To answer my first question, about why they would ever do this, is because today’s browsers are built on age old technologies that can’t cope with today’s online webspace. Let’s backtrack a little. Gears (formerly Google Gears) was created to help cope with the many complaints of having to be online to use web-based applications such as Google Docs or Zoho Writer. So the engineers at Google sought out to create some plugins for today’s browsers to handle and take the load off with the creation of Gears. But there was still a problem, today’s browsers were still extremely limiting and still very crash prone; a browser crash in today’s world could mean loss of much productivity. There was an inability to support multi-threaded applications, innate to all of today’s browsers being steep in old browser tech of static webpages. So Google one-upped itself and thought a few steps ahead and decided to create a browser that not only allowed for multi-threaded applications, but to support multi-processes.

The idea behind Chrome is basic to Software Engineering, instead of handling bugs by simply terminating the process and then restarting, there will be graceful degradation. When browsing and constantly opening up tabs, then closing some, then opening more, today’s current browsers leave fragments and remains in our current OS’s memory, never really clearing them up, so when in today’s world of AJAX powered webpages and having applications and never closing a browser is a must Chrome comes into the picture. Closing tabs means gracefully killing a process, completely removing it from memory, never leaving those fragments in memory to build up and slowing down your machine.

Development
The development of Chrome will be very unique. Google can test Chrome on its massive array web pages using its web crawlers, providing crucial data from millions of websites. Testing has been automated to a level never seen, of course, the usually external testing will still be there, but here is one way testing can come on a massive scale, making bug detection considerably easier.

Interestingly enough, Google Chrome will be a webkit based browser. Most likely because of how Android’s browser is also webkit-based, as well as, the growing popularity of that other webkit based browser made by Apple, Safari and its runaway mobile success. Other additions and modifications to take note is that Google is using a dev team from a company called V8, a company in Denmark, to develop a Javascript Virtual Machine to be able to run multitudes of types of code and render it in the engine. This will be created from the ground up. V8 has taken a drastically new approach to a Javascript VM. Firstly, while Javascript is innately classless, you can still create objects and such and give it new properties on the fly. So with the new Javascript VM, the optimization will be coupled and objects with same properties will be applied the same optimization techniques. The Javascript VM will also optimize on speed, where other Javascript rendering engines run most code on the fly, the code will be offset and the Javascript code will be converted to machine code, which is then compiled and ran on the user’s cpu, similar to Java. V8’s new rendering engine will also optimize on garbage collection, where other engines use conservative garbage collection to guess which objects need to be trashed in memory, V8’s will be able to detect what should and shouldn’t be there so memory is allocated accordingly. The interesting part of this is that, the V8 engine is wholly independent of Google Chrome, so Google wants other developers to be able to use V8’s new rendering browser, “setting a new bar,” as Google calls it.

User Interface
So this tech is all good and all, but what will it look like. Google Chrome’s approach mirror many modern day browsers, multiple detachable tabs, but unlike the other browsers the tabs are outside of the window (think file folders) and the Google’s own take on the Awesome Bar, which they are calling the Omnibox. Unlike Firefox’s Awesome bar, where anything from your web history pops up as you type for inline auto-completions, Google Chrome only auto-completes what you have typed, which can be a good or bad thing. Google Chrome will also integrate search on the fly into their Omnibox, when you have used say Amazon to search for that book of Moby Dick you’ve been itching to read, it automatically remembers it, so which search engines you have used and adds it in. Google Chrome’s startpage takes a varied approach similar in style to Opera’s Speed Dial, but a bit more dynamic. The startpage instead of showing something boring like a blank page, will present to you thumbnails of your nine most visited pages and most often used search engines just to the right of it.

Google has also taken a Safari-like approach to privacy, giving you the option to anonymize your search with a “privacy mode” that switches to a read-only mode when activated, giving you full access to your bookmarks on the web, but never recording anything. Google Chrome will feature its very own pop-up blocker. Google Chrome has also strived to differentiate its users web-experience, by separating as much of the browser from the webapps as possible, hence the name Chrome.

Security
Security in today’s world in most serious. Browser exploits come everyday, Google rationalizes and believes that you can’t ever catch 100% of the malware on the web, so they’ve built a sandbox mode right into Google Chrome, where when activated, webpages submit their right to read/write to your hard disk. Almost every single item is sandboxed, except plugins. Phishing lists are also created and updated continuously on Google’s end and served to you with these updates, so saving you the time to double check if that checking website you used was legit or not. Malicious sites are warned before blacklisted, so exploits can be fixed before your site is on the blacklist.

Plugins
Full access will still be given to plugin developers to fully utilize all of Google Chrome’s features without the fear of never being known. Plugins will run all in their own process, so when a tab crashes, the plugin is still alive and the page is still effectively sandboxed. There will also be full Gears implementation within Google Chrome.

So what does all this mean to you and everybody else. Google just threw down its gauntlet to everyone, Mozilla, Microsoft, Apple and the Opera team and they have open sourced it. Increasing interest substantially as well as opening the floodgates to the millions of developers out there. Just when I thought the race was getting boring, Google has just made it a thousand times more interesting.

Unique as ever Google has enlisted author of Understanding Comics, Scott McCloud to create a web comic

Full Comic Explanation Here by Scott McCloud
Great Over Here From Blogoscoped

--written by Peter To--