With Internet Explorer 9, Microsoft showed Wednesday it's trying to retake the browser initiative.
IE remains the Net's dominant browser. But perversely, it became something of a technology underdog after Microsoft vanquished Netscape in the browser wars of the 1990s and scaled back its browser effort.
That left an opportunity for rivals to blossom--most notably Firefox, which now is used by a quarter of Web surfers, but also Apple's Safari, which now runs on Windows as well as Mac OS X, and Google's Chrome, which aims to make the Web faster and a better foundation for applications.
Microsoft has been pouring resources back into the IE effort, though, and at its Professional Developers Conference in Los Angeles, some fruits of that labor were on display. In particular, Windows unit president Steven Sinofsky showed off IE 9's new hardware-accelerated text and graphics.
The acceleration feature takes advantage of hitherto untapped computing power in a way that's more useful than other browser-boosting technology--Google's Native Client to directly employ PC's processor and Mozilla's WebGL for accelerated 3D graphics, for example--according to Dean Hachamovitch, general manager of Internet Explorer.
"This is a direct improvement to everybody's usage of the Web on a daily basis," Hachamovitch said in an interview after Sinofsky's speech. "Web developers are doing what they did before, only now they can tap directly into a PC's graphics hardware to make their text work better and graphics work better."
... Read moreWith a project called Closure Tools, Google plans on Thursday to start helping developers who aspire to match the company's proficiency in creating Web sites and Web applications.
Google is a strong proponent of using JavaScript to write Web-based programs, part of its Web-centric ethos. Indeed, the company has pushed the language to its limits with services such as Gmail and Google Docs, and it developed its Chrome browser in part to enable JavaScript programs to run faster.
But writing, debugging, and optimizing heavy-duty JavaScript can be difficult--in part because a given JavaScript program sometimes works differently on different browsers. Google's open-source Closure Tools project is an attempt to help with some of these challenges.
The first in the suite of tools is the Closure Compiler, a software package designed to boil down a JavaScript program so it's smaller and runs faster. For example, a function named DisplayAddress() could be replaced with just a().
Along with the compiler come some extra tools that run in the Firefox browser. One, Closure Inspector, is an extension for Firefox's Firebug add-on designed to help programmers understand and debug the rewritten JavaScript--linking a() back to DisplayAddress(), for example. Another add-on for the Google Page Speed extension lets programmers see how much the compiler helped.
Google also plans to make the compiler available as a Web application hosted on its Google App Engine service.
The second element is called the Closure Library, a collection of prebuilt JavaScript code that lets programmers handle relatively sophisticated technology--arrays and string manipulation, for example.
Last are Closure Templates, more prewritten code to ease creation of JavaScript and HTML user interfaces.
In an earlier era, programming tools were expensive packages bought by a select few, but open-source software, new marketing strategies, and new business methods have made that approach the exception rather than the rule these days. Now programming tools are often a means to another end--encouraging programmers to produce the software that will make Windows or the Palm Pre useful and therefore popular, for example.
In Google's case, the objective is often to make the Web more popular because it sees more activity on the Web as corresponding directly with more activity on its revenue-generating search site. Among the high-profile projects to this end are Chrome, Chrome OS, and Android, all subsidized by Google's powerful search-advertising business.
One interesting contrast to Closure is another Google project called Google Web Toolkit. It's designed to accomplish some of the same goals as Closure, including paving over browser incompatibilities and producing high-performance JavaScript. But with GWT, coders write programs in Java that gets translated into JavaScript.
So one last question: why the name?
Google's reply: "Being a functional language, the concept of a function closure is fundamental to the JavaScript language."
(Credit:
Snac)
Not everyone has a smartphone, and strange as it may sound for those who can't live without their mini call-placing computers, avoiding an $80 to $100 monthly data plan is often driving that conscious choice. But it is possible to get a data add-on to a cell phone plan for as little as $10 a month. For that set of in-betweeners, a program like Snac could bridge the gap between a modest data plan and accessing Web content like social network feeds and news stories.
Snac, in open beta, provides a dashboard of thumbnail-size widgets that periodically grab content from Facebook, Twitter, Gmail (POP3), the weather, Google Calendar, news sources, and so on. Snac precaches a few layers of content, so that when you click a widget, you'll be able to immediately see headlines, status updates, and so on, on a small pop-up window.
Snac isn't just read-only either. For instance, you're able to update your social networking status and add comments in Facebook. Ditto with Twitter, but Snac isn't equipped for direct replies. Opening a news story pushes you to your phone's default browser, but Snac CEO Mark Caron tells us they'll soon add an in-app browser for reading purposes. At this point, you're unable to configure how often Snac pulls in fresh content, but Caron says they're looking into that, too.
Caron demoed Snac beta on a Sony Ericsson TM 506 phone, and a private beta version for BlackBerry. It works on almost 300 Java feature phones, but also on Symbian and Windows phones. There's no limit to how many widgets you can install from the Snac gallery, and managing the dashboard is straightforward from the phone settings and from the Web.
The free, ad-supported app isn't much of a looker, with a boring background, blocky typeface, and sketchbook widget design. But it worked quickly and well in our demo. We'll be keeping an eye out for future developments.
Google released an Internet Explorer plug-in Tuesday designed to let Microsoft's browser use the features and performance of Google's own Chrome browser.
The software, called Google Chrome Frame, lets IE 6, 7, or 8 use Chrome to render Web pages and execute their JavaScript programs, Google said. To use it, people must install the open-source plug-in, currently in the developer preview stage, and Web developers must insert a line of code onto their Web sites that engages Chrome Frame when a person visits the site.
"For users, installing Google Chrome Frame will allow them to seamlessly enjoy modern Web apps at blazing speeds, through the familiar interface of the version of IE that they are currently using," said Google programmer Alex Russell and product manager Mike Smith in a blog post.
But the plug-in might needle its rival more than revolutionize Web browsing. For one thing, it takes a long time to get a lot of Web developers to update their sites. For another, how many people dissatisfied with IE's performance haven't already installed a higher-powered browser?
Google argues that the feature will appeal to some folks, though, including people in corporate settings who might not have a choice of browser and people who prefer IE's interface, said spokesman Eitan Bencuya. And people are familiar with plug-ins as a way to expand what browsers can do.
"It's a much lower barrier to entry than switching browsers," Bencuya said.
He added that Google has built support for the feature into one of its own Web sites, the Google Wave project that's a hybrid of e-mail, instant messaging, and wiki collaboration.
These days, Mozilla's Fennec and the Skyfire browser have been stealing all the thunder in the mobile browsing space. On Wednesday morning (that's Tuesday night for us in San Francisco), Opera yanked some of it back with the release of Opera Mini 5 beta for Java phones.
Introducing a graphically enriched layout topside and new features below decks, the new Opera Mini beta browser is snappier, more attractive, and more advanced than last year's predecessor, Opera Mini 4.2. Mini 5 beta brings over several features from Opera's desktop browser (Opera 10 for Windows | Mac.) Tabbed browsing is among them, as is a password manager. Each page opens with Speed Dial, a grid of nine thumbnail images and Opera Desktop mainstay, that you assign to favorite Web sites and can select among to quickly launch a Web page. The Speed Dial view replaces Opera Mini's previous landing page, a tangle of links capped with a search box and URL field. These thumbnail images make the landing page more meaningful, both in giving users a visual they can instantly recognize, and creating an easier target for users to accurately hit on touchscreen phones than a scrawny little link.
While the URL field and search bars haven't joined together in this beta as they have in other mobile browsers and in most desktop browsers out there, Opera has at least consolidated the two onto a single line. To address another long-overdue fix, Opera now lets you type directly into a text field. In previous versions, clicking a field opened up a blank page, where you were prompted to start typing before you could return to the main interface.
Opera Mini gets into Opera Desktop's Speed Dial start screen.
(Credit: Opera Software)Opera Mini's navigation menu received another overhaul in Mini 5 beta. Opera moved it up to the top and made it completely icon-based. Press downward (on a D-pad for a keypad phone) to engage more items, like bookmarks, history, settings, and the Find in Page search tool, a new one for Opera Mini. Find in Page has previously been available in Opera Mini; it's nice to see it return.
The password manager that's new to Opera Mini works as expected, producing a dialog box the first time you log into a site asking if you'd like it to remember your credentials. You can turn this off in the Privacy portion of the Settings submenu.
Many additional features carry over from previous Opera Mini versions, including options to view the page as you would from the desktop versus a mobile view. There are also the usual shortcut keys and support for landscape mode on most phones (not on BlackBerrys, unfortunately, an ongoing omission). There are also additional options that pop up in response to long presses on the 'select' key or on the touchscreen, like for selecting and copying text, opening the image, and now, for opening content in a new tab.
... Read moreOpera Software has completed its first release candidate of Opera 10, a browser that the company says has better performance, a Turbo mode for slow Internet connections, support for a variety of Web standards such as Web fonts, and improvements to the Opera Mail feature.
"Now, we are very close to releasing the best browser in Opera's long history," Jan Standal, Opera's vice president of desktop products, said in a statement. "We hope everyone who has helped us test our browser thus far will put the release candidate through its paces."
The new Carakan JavaScript engine, which is used to run Web-based applications such as Google Docs, isn't done yet.
"It won't be ready for (Opera) 10 final, but rest assured that it will be impressive when it comes," spokesman Thomas Ford said. He said Opera won't comment on the timing of the new engine's release until it enters alpha testing.
Firefox, Safari, and Chrome also all are working furiously on better JavaScript performance too, in an effort to make the Web a better foundation for applications.
The new Opera release candidate is available for download for Windows, Mac, and Linux.
Opera has been available for years as an alternative to the dominant Microsoft Internet Explorer, the second-ranked Firefox, and Apple's Safari. It was pushed into fifth place with the arrival of Google Chrome. The Opera browser often charts new territory, though. For example, its Speed Dial feature, which presents an array of Web site thumbnails when a person opens a new browser tab, was first introduced in 2007. A similar feature can now be found in Chrome and Safari, and Firefox may add something comparable.
Mozilla has released the first alpha version of Firefox 3.6 for Windows, Mac, and Linux, a browser with speed improvements and new features the organization hopes to finalize faster than its predecessor.
"Unlike the year that passed between Firefox 3 and Firefox 3.5, we expect that this 3.6 release will be released in a small number of months," Mozilla evangelist Chris Blizzard said in a blog post Friday.
Firefox 3.6, code-named Namoroka, has a variety of changes, but it's not as dramatic a departure as 3.5 was from 3.0. Among the 3.6 features are faster JavaScript, the Web programming language Firefox executes with its TraceMonkey engine; faster page-rendering speed; some new features for CSS (Cascading Style Sheets) technology for controlling some of the look of a Web site; and a feature called the compositor that handles complicated layout circumstances better.
Performance is a big issue with browsers these days as people spend more time using them and programmers create more sophisticated sites and applications that live on the Web. All major browser makers are emphasizing performance improvements in their newest versions.
Download links for the first Firefox 3.6 alpha are at the Mozilla Developer Center.
Satisfied that its security underpinnings are solid, Google has promoted its open-source Native Client technology to accelerate Web applications out of its research phase and is taking steps to build it into the Chrome Web browser.
"Based on our experience to date, we believe that the basic architecture of our system is sound and the implementation is supportable. So now we are undertaking a number of tasks to transition Native Client from a research technology to a development platform," said Brad Chen, Google's Native Client engineering manager, in a mailing list announcement Wednesday.
Brad Chen, engineering manager of the Google Native Client
(Credit: Stephen Shankland/CNET)Native Client, called NaCl for short, is a mechanism to run software downloaded over the Web directly on x86 processors such as Intel's Core line. The key motivation is to attain the speed of regular "native" software installed on a computer rather than the much slower JavaScript environment that sophisticated Web sites use today. It's one part of Google's broad effort to evolve the Web from a collection of relatively static sites into foundation for more powerful applications.
Executing native code from the Web is easy--until you start trying to worry about security risks. To this end, Native Client examines software before it runs to block software that takes a variety of prohibited actions, an idea called static analysis, and it runs the software in a protected sandbox.
"We recognized the underlying technology to be ambitious and risky, and felt strongly than a generous measure of public scrutiny was appropriate before we committed to any definite plans," Chen said. Satisfied that Native Client passed muster, Google will remove various security constraints such as the inability to execute Native Client software downloaded from the open Internet, he said.
Native Client was first introduced in December a browser plug-in, but Google doesn't like that approach.
"We recognize that there is well-justified resistance to installing browser plug-ins. For this reason we have a strong preference for delivering Native Client pre-installed or built into the browser, and we'll be focusing on that as our main strategy for delivering Native Client to users," Chen said.
And now we see one reason why Google is interested having a browser of its own available: "Careful readers may have already noticed evidence of integration into Chromium in the Native Client source," Chen said, referring to the open-source project that underlies the Chrome browser.
Google touted Native Client at its Google I/O conference in May, showing off a Web-based photo editor as an example of the processing power the technology offers. Google also is trying to pair Native Client with another company project, O3D, which lets browsers take advantage of hardware to accelerate 3D graphics.
Yesterday's introduction of Chrome version 2.0.172.28 was touted by Google as being up to 30 percent faster for handling JavaScript. After using the update to the stable release extensively for the past day and running it through two JavaScript tests on two computers, it's conclusively faster than the previous stable version of Chrome.
In addition to being faster, the new Chrome now has a full screen mode.
(Credit: Screenshot by Seth Rosenblatt/CNET)The two computers I used were a Windows Vista Service Pack 1 desktop with a Pentium 4 processor running at 3.00 GHz and 2 GB of RAM, and a Windows XP Pro Service Pack 3 laptop with a Core Duo T9400 processor running at 2.53 GHz and 3 GB of RAM. Chrome was benchmarked by the Webkit test, SunSpider, and the JavaScript-only sections of the Mozilla test Dromaeo. Chrome was tested with no other tabs open and no other programs running on the computer.
When testing SunSpider, Chrome v1.0.154.65 scored 919.2ms on the laptop, and 1864.2ms on the desktop. Chrome v2.0.172.28 scored 583.6ms on the laptop, and 1323.4ms on the desktop. The laptop score was 36.6 percent faster, and the desktop was 29 percent faster.
Running both versions of Chrome through Dromaeo came up with similar results. Keeping in mind that the higher number is better for Dromaeo's tests, the older Chrome scored an overall 113.25 runs/s on the laptop. The newer one hit 139.90 runs/s, an improvement of 23.5 percent. The desktop results were the inverse of the laptop's.
Where the SunSpider results showed greater gains for the new Chrome on the laptop, the Dromaeo desktop tests showed an improvement of 33.8 percent. Chrome v1.0.154.65 hit 146.63 runs/s while v2.0.172.28 scored 196.29 runs/s.
Also new: Users can set forms to autofill in the Options menu (foreground, right), and selectively remove thumbnails on the New Tab landing page (background, left).
(Credit: Screenshot by Seth Rosenblatt/CNET)Actual results will vary based on your hardware and other program running simultaneously, but it's irrefutable that the new version of Chrome, with its updates to the WebKit rendering engine and Google's V8 JavaScript engine, is significantly faster than its predecessor. I strongly recommend that you upgrade if you've been using Chrome for script-intensive Web apps.
Also new in this version of Chrome is the F11 hot key to toggle full screen mode, which drops all hints of the browser frame except for the scroll bars--but only if they're appearing on that site in standard mode. There's also a new form autofill under Options in the Tools menu, and users can selectively remove thumbnails from the New Tab landing page.
Google has shifted the JavaScript engine that powers its Chrome Web browser into a higher gear.
The company announced Thursday that an update to Chrome's V8 JavaScript engine, combined with a new version of the Webkit browser engine, should improve the loading speed of JavaScript-heavy Web pages by up to 30 percent. The updates will be automatically downloaded to existing copies of Chrome.
JavaScript engines are one of the new fronts in the browser wars, with various vendors touting the performance of their browsers this year in hopes of unseating the competition. Chrome did very well on CNET JavaScript tests earlier this year, besting Firefox, Safari, and Internet Explorer.
Google also announced a few other changes to Chrome, including the addition of new features that let you erase embarrassing (or NSFW) Web sites from the most-visited list that appears when you open a new tab in Chrome. And Chrome now has a feature found in many browsers: form autofill.





