While Firefox 3 pushes the boundaries of a browser, delivering a cleaner graphical interface, numerous security upgrades, and a variety of new features, it still is far from the "perfect browser".  Of course no browser on the market could earn such a distinction, be it Opera's "Kestrel" 9.5 browser, which is shaping up to be very nicely, Microsoft's Internet Explorer 8, which is sure to deliver top end performance, or Apple's Safari v3.1 browser, which although plagued with bugs of late is still one of the fastest, most compatible, browsers around.

However, perfection is exactly what Mozilla wants for its Firefox 4 browser and it intends to get it through outside the box thinking.  With the release of Firefox 3 imminent, Chris Beard, VP of Labs for Mozilla, has started to talk Firefox 4.  Beard is working on many features which, while their underlying components might start to pop up in Firefox 3, won't truly see the light of day until Firefox 4.

Beard's basic push is to un-isolate the browser.  He says that after 10 years the browser still remains isolated from your browser on other machines and from your computer environment. This leads to the focal point of Mozilla's efforts for Firefox 4.  Mozilla is pushing strongly for two very different new lines of research:  Prism and Weave.

Prism is the main path of Mozilla's efforts to make the browser into almost a virtual OS, with applications, workspaces, and more advanced resource management and graphical abilities.  Mozilla says the HTML 5 and Prism will make Google Gears obsolete.  Google Gears is a beta service from the search giant which offers a way of accessing its online tools and applications, offline via clever caching and scripting.  Mozilla also claims Prism will similarly be a death knell for Microsoft Silverlight and Adobe Air, two programs similar to Google Gears, attempting to meld online and offline applications together.

Mozilla's goal is to be able to take any website or application and turn it into an application that can run directly from the desktop.  While part of the emphasis is removing the classic need to go into a browser to access these applications, another focus is to make the applications available when you are offline.  HTML 5, the upcoming next standard of the classic internet language will be updated to provide explicit support for offline/local resources, which should significantly aid Firefox 4 towards accomplishing these ends. 

While little of Prism will appear in Firefox 3, Weave will see some of its underpinnings crop up in Firefox 3.  Weave moves to expand Firefox in the opposite direction as Prism -- instead of expanding towards the desktop, this one looks to expand the internet boundaries of what Firefox can accomplish.  The key emphasis of this drive is to provide a way of syncing a user's online preference whenever they use a browser anywhere.  Among the elements of this are syncing bookmarks, home pages, favorites, and passwords.  With the increased mobility from improved internet on mobile phones, consoles, and other devices, this becomes a particularly useful goal.

Firefox 3 implements the skeleton of this, with a new transactional database for store bookmarks and favorites.  In Firefox 3 this simply allows more efficient and unique management of these items, but does not offer inter-browser syncing.  Beard hopes to have the extensions necessary to provide syncing ready to roll out for Firefox 4.  Firefox 3 will implement some information being fed into the browser when online, which should pave the way for this.  In Firefox 3, streaming updates to malware signatures will help shore up security by offering protection against new phishing and drive-by download scams.

Aside from Weave or Prism specific efforts, Beard is trying to focus on making the browser take effort off the users shoulders by helping them figure out web addresses.  Firefox 3 implements a basic version of this called the "awesome bar", the new smart web-address bar.  Most users of Firefox 3's betas should already have noticed this feature.  Beard wants to go beyond this and allow for a basic "linguistic user interface" in Firefox 4 which will allow users to type basic commands in plain English.  He cites Enso and Quicksilver as to utilities that implement functionality similar to what he hopes to offer.

Aside from these core improvements, Beard warns that there may be some other "crazy ideas" in store for Firefox 4.  With such ambitious goals, it should be interesting to see what Beard and his team can put together for 2009-2010.