Webware

Mozilla won't make a 2009 deadline for releasing Firefox 3.6 and is giving itself more time to complete a major update, version 4.0.

The organization behind the open-source Web browser had predicted a final release of Firefox 3.6 in December 2009, but the Mozilla Web site now includes "ship Firefox 3.6" as a goal for the first quarter of 2010.

In addition, Firefox 4.0, which had been due in 2010, now is "aimed at late 2010 or early 2011," with a beta due in the summer of 2010, according to Mozilla.

Schedule delays are common in the software world, but browser development is furious these days with the arrival of Google's Chrome into the market, Apple helping to expand the frontiers of what the browser can do, Opera trying to dramatically speed up JavaScript execution and display performance, and Microsoft getting more ambitious again with Internet Explorer. "We've always been more quality-driven than time-driven, but we understand timing in the market matters to our users and our competitiveness," said Mike Shaver, Mozilla's vice president of engineering, in an October interview.

... Read more

If you want to see the scale of browser makers' ambition to remake not just the Web but computing itself, look no farther than a new 3D technology called WebGL.

The WebGL vision is simple. You're running around in a video game universe, blasting radioactive aliens--but you got there by visiting a Web site, not by installing the game on your PC.

This sort of computationally demanding chore contrasts sharply to with today's Web, whose top-notch programmers strain to reproduce bare-bones versions of the rich capabilities open to applications running natively on a computer.

WebGL, while only a nascent attempt to catch up, is real. WebGL now is a draft standard for bringing hardware-accelerated 3D graphics to the Web. It got its start with Firefox backer Mozilla and the Khronos Group, which oversees the OpenGL graphics interface, but now the programmers behind browsers from Apple, Google, and Opera Software are also involved.

Perhaps more significant than formal standards work, though, is WebGL support in three precursors of today's browsers--Minefield for Mozilla's Firefox, WebKit for Apple's Safari, and Chromium for Google's Chrome. Opera has started implementing WebGL, too, said Tim Johansson, Opera's lead graphics developer.

With a little tinkering--check the instructions and caveats below--you can give it a whirl, too. Overall, I was favorably impressed with the technology.

CNET News Poll

Will you use WebGL?
Browser makers are building 3D technology into their products. Will you use it?

Yes, bring on the 3D Web
I'll stick with Flash graphics
Skip it. Direct3D is the way to go
Google's O3D looks better
Yuck. More spinning cubes?

View results

Its performance certainly isn't enough for a competitive first-person shooter, but it's approaching utility for casual gaming. And because of how WebGL elements can be integrated with the rest of a Web site's code, it's got some advantages.

What is WebGL?
WebGL is one of a handful of efforts under way to boost the processing power available to Web applications. It marries two existing technologies.

First is JavaScript, the programming language widely used to give Web pages intelligence and interactivity. Although JavaScript performance is improving relatively quickly these days in many browsers, programs written in the language are relatively pokey and limited compared with those that run natively on a computer.

... Read more

Google's browser has passed Safari in terms of worldwide browser usage--at least by one measurement.

NetApplications' measurements of browser usage share, which track which browsers individuals use based on visits to the company's network of Web sites, gave Chrome the third-place spot after No. 1 Internet Explorer and No. 2 Firefox for the week of December 6 through 12, according to a Computerworld story Tuesday. Chrome had 4.4 percent share to Safari's 4.37 percent.

Google released beta versions of Chrome for Mac OS X and Linux on December 8. Earlier, only developer channel versions had been available. Google plans to release the "stable" versions January 12, according to the Chromium development calendar.

Take these usage share numbers with a grain of salt. Even though 0.03 percentage points still is a lot of people in the real world, it is a small fraction, and a change in Net Applications' assumptions in August led to share changes two orders of magnitude more dramatic. Weekly statistics also vary: Although Firefox cleared 25 percent share in one week of November, it averaged only 24.72 percent for the overall month.

I've asked various browser makers about how trustworthy they view NetApplications' statistics to be. The answers generally are favorable but not ringing endorsements.

Regardless of the precise details, though, the Chrome trajectory is upward: its November usage share was 3.93 percent to Safari's 4.36 percent.

And although Google relied on word of mouth for promoting its original online search product, it's taking a more active role with Chrome. The latest example: a "Chrome for Christmas" site that lets people send invitations to download Chrome.

Firefox proved that a browser not bundled with an operating system can be successful, and Chrome could show the idea isn't a fluke if its growth continues.

As a company that has built a business model atop trust, Google is in a sticky position as it prepares to formally introduce the Nexus One phone.

Google employees were given free Nexus One phones at a company party Friday night, and the Internet went into a tizzy. Reports surfaced later in the weekend that this device was the long-awaited Google phone, the company's answer to Apple's strategy of controlling the hardware, software, and distribution model with the iPhone, rather than the partner-oriented strategy of developing the guts of the operating system and letting partners each put their own stamp on the finished product.

Just two months ago, Google's Andy Rubin rolled his eyes when asked about an analyst report picked up by TheStreet.com that said Google planned to pursue this exact strategy. He said Google had no plans to make its own hardware--which is one thing since smartphones are almost exclusively manufactured by contractors in China and Taiwan--but he took a further step in spending about 10 minutes arguing why it would be a bad idea for Google to design its own phone and sell it outside of carrier channels.

That line of thinking resonated with many who follow Google and the mobile industry. After all, Google's stated goal for Android ever since the project was revealed in November 2007 was to create an "ecosystem" of multiple phones that would help improve access to the mobile Internet. And Google seemed to finally reach that goal this year, with over a dozen phones in the wild and more promised from some of the world's leading phone makers and wireless carriers.

But if the reports are correct, Google is about to make a radical departure from that strategy. And Google's new course would take it down a path that could sow distrust among the company's Open Handset Alliance partners, who must now be wondering if they're about to get into a marketing war with one of the tech industry's richest companies.

Katie Watson, a Google representative, said on Sunday that the company has confirmed nothing about its plans for the Nexus One, described as a "dogfooding" experiment for internal testing by the company in a blog post Saturday.

In the rush to anoint the Nexus One as the Google Phone, it's quite possible that the tech industry glossed over the fact that Google already sells Android phones, albeit on a limited basis. For quite some time, registered Android developers have been able to buy completely unlocked versions of the G1 and the T-Mobile MyTouch3G (also known as the Google Ion) for $399.

So there is a solid chance that the Nexus One is merely the Android Dev Phone 3, following the Dev Phone 1 (G1) and Dev Phone 2 (MyTouch or Ion). Just this year, Google handed out Dev Phone 2 models branded as the Google Ion to attendees at Google I/O 2009, but if regular people want to buy that particular phone they have to get the MyTouch3G from T-Mobile with a two-year contract.

It does seem clear that Google has played the premier role in designing the software for the Nexus One. In the company's blog post over the weekend, it said "we recently came up with the concept of a mobile lab, which is a device that combines innovative hardware from a partner with software that runs on Android to experiment with new mobile features and capabilities, and we shared this device with Google employees across the globe."

But the key unconfirmed detail is how Google plans to sell this phone. According to The Wall Street Journal, Google plans to sell this phone unsubsidized on its own, with consumers able to choose a wireless service provider after the fact. However, according to corporate sibling Peter Kafka at All Things D's MediaMemo and Reuters, Google has plans to hook up with longtime mobile partner T-Mobile to help sell the Nexus One through Google's Web site for $199.

How will Google market this phone? Anyone with a television set has likely seen an ad over the last month for the Motorola Droid, an Android phone sold for Verizon's network that has been billed as one of the best Android phones to date. It was also the launch pad for a long-term pact between Google and Verizon that will supposedly produce a family of devices based on Android.

If Google plans to sell the Nexus One directly to consumers, will it insist upon using its brand as the lead brand, rather than the "With Google" branding found on the back of many Android phones? Will it blast the airwaves during the NFL playoffs in January to trumpet the arrival of the Nexus One, perhaps just in time for the Super Bowl? And how will that affect partners such as Motorola and Verizon that have sunk so much money into promoting the Droid, only to see rumors of a Google Phone leak out at the worst possible time: the height of the holiday shopping season?

This could be a very telling moment in Google's history. At the moment, Google's mobile division does not seem to be completely in control of the message it wants to send consumers, partners, and competitors.

If Google really does plan to sell the Nexus One directly to consumers and compete with its customers, it has chosen an interesting way to announce it to the world, keeping the Google Phone rumor mill alive for months while publicly denying such plans. Apple has employed such a marketing strategy for years, insisting on near-silence regarding future product plans but benefiting enormously from the frenzy of interest in every little morsel that mysteriously pops up regarding those plans.

However, Google is not Apple. Google public-relations representatives will sheepishly admit that they have little control over how Google rolls out its products: Google is a company run by engineers, and engineers push the button when the product is ready to ship.

But when you're working in an environment with multiple partners that have competing interests, any confusion over your future plans--especially plans that would appear to yank the floor away--can breed distrust among those partners. One of Google's largest problems right now is that it has built a business model geared around the notion that it can be trusted with almost unprecedented control over the flow of information across the globe, and any cracks in that wall of trust will be exploited by its enemies.

With the way details have trickled out about the Nexus One, Google has either alienated current and future Android partners by muscling in on their turf, or set up thousands of eager smartphone consumers looking for an open alternative to the iPhone for disappointment when they realize Google merely plans to sell an expensive unlocked phone to a limited audience, if at all.

After all, Google essentially declared in its blog post that employees are testing a product with "new mobile features and capabilities" that presumably can't be found on the current crop of phones. It's almost the same language Google used to introduce Chrome OS ("our attempt to re-think what operating systems should be") while insisting that it had no competitive reasons for introducing that Netbook operating system.

Few believed that line with Chrome OS, and fewer still will believe that Google is creating Android for the betterment of humanity if it really plans to sell its own phone.

What's behind the Google Chrome OS, technologically and from a business perspective? This week on the Roundtable, I discuss the pending operating system with CNET writers Stephen Shankland (Deep Tech) and Gordon Haff (Pervasive Data Center).

Listen now: Download today's podcast

Subscribe with iTunes (audio)
Subscribe with iTunes (video)
Subscribe with RSS (audio)
Subscribe with RSS (video)

Reporters' Roundtable #13: Google Chrome OS ... Read more

Mozilla Messaging pushed the stable release of Thunderbird 3 out of the nest on Tuesday, and there's a lot to like in case you haven't been following the beta development of this Outlook alternative. The long-overdue Thunderbird 3 is available for Windows, Mac, and Linux, and it introduces several hefty new features and some long-needed improvements, including an overhauled search and message indexing, tab support, and a revamped setup wizard that's designed to make new account setup quick and painless.

One feature that isn't included is the calendaring add-on, Lightning. Originally, Mozilla had planned to bake the extension into the program, but decided back in February 2009 to change course and leave it up to users to download. Although Thunderbird natively comes with Microsoft Exchange support, there's no calendar and therefore no meeting support in the default Thunderbird installation. Along with Lightning, there's an essential Google Calendar add-on for Lightning that gives Google users calendar support in Lightning. Currently, the only version of Lightning that works in Thunderbird 3 is the nightly build, available here.

Even without Lightning, Thunderbird makes for an excellent desktop-based e-mail client. Beyond Outlook replacement, it also makes a savvy offline or local-storage tool for the various Web mail providers. Gmail integration has existed in Thunderbird for a while, but improvements in version 3 include better recognition and integration of Gmail's special folders. These include Sent and Trash, and the non-English versions of Gmail. The All Mail option in Gmail defaults in Thunderbird to the Archives folder.

Undeniably, the killer feature in Thunderbird 3 is the search. The most obvious competitor, Microsoft Outlook, doesn't offer anything that comes close to the level of granular control that Mozilla has given Thunderbird users. The new search bar is dominant at the top of the interface, and is set by default to search all messages. When you search, a new tab will open with your results organized as shown in the screenshot above. Filters based on e-mail addresses, folders, and tags appear on the left, while the majority of the window is given over to displaying summaries of the e-mails that meet your criteria. There's also a timeline bar graph at the top of the results. Click it, and then mouse over any of the subcategories to see how they occurred over time.

You can change the search box to one of several filters, including Subject, From, Recipient, To, CC, and Message Body. Frustratingly, you can't filter by Tag. You can also save any of these filtered searches as a virtual folder. Editors' note: The previous two paragraphs have been rewritten for clarity.

There is one drawback to the search: the first time it indexes your messages, you're potentially in for a long, long wait. In testing, this depended entirely on the number of messages in your folders. High volume accounts, whether locally archived or all on a server, should probably set their indexing to run overnight. After the first indexing, each new e-mail is added as it comes in.

E-mails open by default into new tabs, making the e-mail reading experience far more similar to the Web-browsing one. This can be toggled under the Advanced section of the Options, under the Reading and Display tab. The hot keys for the e-mail tabs have been mapped the same as in Firefox, so middle-click an e-mail to open it in a new tab but retain your focus on the current tab. The CTRL+Tab hot key combo will cycle through your tabs, and there's an open tab button on the right side of the tab bar to help manage your tabs.

There's a new activity manager that records all interactions between your e-mail provider and Thunderbird, making it easier to track down errors when you send or receive mail. There's also an entirely new system for archiving messages based on Gmail's "archive and forget it" method. The new version offers the traditional multiple-folder-based solution, as well as the new dumping-ground style, which can be activated via the "A" hot key. Thunderbird 3 supports Firefox personas, too, further reinforcing their shared architecture.

Other changes include major code improvements. The setup wizard now looks to mozillamessaging.com for additional information on how to configure the account. This changes how new Web mail accounts are created. Mozilla has said that only the domain name from your e-mail address gets sent to Mozilla's servers, and that the entire process falls under the Mozilla's privacy policy. Nevertheless, it's a move that's likely to cause some concern among privacy advocates.

The compact header mode has been deleted, which is sure to annoy those who like using Thunderbird on smaller-form computers like Netbooks. Windows users should see Thunderbird results appearing in federated searches in Windows Vista and Windows 7, while Mac users will find Growl notification support for new e-mails, integration with Spotlight and the Mac OS X address book, and support for Mail.app. The full changelog for Thunderbird 3 can be read here.

Thunderbird 3 rates as a top-notch e-mail client, and it's definitely the best freeware one around. It will require some fidgeting to get it to be usable in a corporate environment, but it's far more scalable to user needs than anything else currently available.

Two key pieces of Google's effort to make Chrome a more competitive browser fell into place on Tuesday as Google released beta versions of the browser for Mac OS X and Linux.

Tuesday's software release is a version of Chrome that had previously been available only as developer preview software for Mac and Linux machines. "It took longer than we expected, but we hope the wait was worth it," product manager Brian Rakowski said in a blog post.

Macs are widely used, if not as common as Windows machines, and there's been some demand in tech circles for the Mac version of Chrome. Linux, while less widely used among ordinary computer users, has importance of its own: it's the foundation for Chrome OS. That's the browser-based operating system Google hopes will be popular on Netbooks starting next year.

According to the Chromium development calendar, the beta versions are scheduled to graduate to the next level of maturity, "stable," on January 12. Chrome for Windows graduated out of beta almost exactly a year ago.

Google doesn't emphasize product version numbers in the project, instead automatically delivering updates behind the scenes to the browser that take effect when it's restarted. But it does use version milestones to keep track of development internally.

The biggest new feature of Chrome 4.0 is support for extensions, which let people customize the browser. In the Mozilla world, they're called add-ons, and they've been a big part of Firefox's success.

Extensions aren't useful, though, unless people can find them. Google on Tuesday also launched a Chrome extensions gallery page.

There are more than 300 extensions available for Chrome, extensions programmers Aaron Boodman and Erik Kay said in a blog post.

However, extensions on the Mac aren't yet available, though they had been for a time in the developer-preview version. "Extensions aren't quite beta-quality on Mac yet, but you will be able to preview them on a developer channel soon," Rakowski said.

Also on the Chrome for Mac to-do list: a bookmark manager, PDF viewing in the browser, bookmark synchronization, 64-bit support, and my personal favorite differentiator of Firefox 3.6 on the Mac, full-screen support.

Why try Chrome?
For those of you new to Chrome, here's a brief version of why it's my default browser on both Windows and, as of about a month ago, Mac OS X. Your preferences and needs may vary, of course, and I still use Firefox every day, too.

• Speed. It's fast to start up, though not quite as snappy as it once was now that it's not so bare-bones, and rivals are making progress. It's also fast loading Web pages and running JavaScript programs on them.

• Tabs. I spawn innumerable new tabs all day long, and when it takes a long time (I'm looking at you, Internet Explorer), I get infuriated. I also like the order in which new tabs arrive, a style Firefox is mimicking.

• The omnibox. It's a single bar that merges the utility of an address bar and search bar. I hit Ctrl-L (on Windows) or Command-L (on Mac) to pop my cursor up there, and start typing. One nice--if somewhat obscure--feature is fast site search on some domains, so for example I can type A, M, tab, and up pops an Amazon.com icon; what I type afterward is entered as a search on Amazon. That conveniently gets me straight to the search results so I don't have to see yet another Kindle ad.

• A minimal user interface. When browsing, I like my user interface to step aside and make way for the Web page. Scrolling was a wonderful innovation in computers a few decades ago, but I like to avoid it when I can. Chrome puts tabs in the real estate ordinarily devoted to a program's title bar and shuffles the menu controls off to the right of that tab strip (though the Mac version gets a regular menu bar).

Another potential perk: avant-garde Web technology, including WebGL and O3D for accelerated 3D graphics and Native Client for speeding up Web apps with direct access to a processor, are being built into Chrome. Another such Google project, Gears, is already built into Chrome--though Gears doesn't work on Mac OS X 10.6.

There are things you might miss--the full panoply of Firefox extensions, toolbars from Google or others, print preview. And the "browser not supported" error messages on various Web pages are annoying, though in my experience there's rarely an actual compatibility problem. Overall, I like it.

Is Google spying on me?
If you're worried about what new data Google will be able to harvest on you, I recommend a close read of Google's Chrome privacy page. This doesn't worry me much, but I may be insufficiently paranoid. In my opinion, the biggest thing is that Google stores 2 percent of the data it gathers when people type text into Chrome's combination search and address bar, called the omnibox.

That means Google can see not only what you're searching for (as it would for any Google search), but what Web site addresses you're typing as well. The data is anonymized within 24 hours, Google said.

Also, Chrome has a feature called DNS pre-fetching that tracks down the Internet server addresses on Web pages in anticipation that you'll be clicking links on the page. So Chrome--and Google, too, if you're using Google Public DNS--retrieves this information from the Internet.

Updated at 12:30 p.m. PST and 1:20 p.m.. Added further detail.

After more than five years as a publicly available test version, Gmail shed its beta label in July. Now one feature key to the Net giant's cloud-computing aspirations, offline access to Gmail, also has grown up less than a year after its debut.

"Offline Gmail is graduating from Labs and becoming a regular part of Gmail," Google programmer Aaron Whyte announced the change Monday in a blog post.

Offline Gmail support, which relies on a Google browser plug-in called Gears, lets people read, search, organize, and compose e-mail even when there's no Net connection; sent messages are queued up in an outbox for delivery when the network access is restored and the account on the computer can resynchronize with the server.

"Offline Gmail has proven particularly useful for business and schools making the switch to Google Apps from traditional desktop mail clients--they're used to being able to access their mail whether or not they're online, and Offline Gmail brings this functionality right to the browser," Whyte said.

Google Apps, a bundle that includes Gmail, Google Calendar, and the Google Docs suite of online applications, is available for free for educational users or smaller organizations. Premiere accounts cost $50 per person per year, and Chief Executive Eric Schmidt called such enterprise-oriented services Google's "next big billion-dollar opportunity."

Gears is built into Google's Chrome browser, but other browsers rely on a plug-in. However, Google has stopped developing Gears in favor HTML5's equivalent features. That overhaul of the standard for displaying Web page includes local data storage on a computer as one feature, and it's now enabled by default in Chrome even though HTML5 isn't a final standard yet.

Updated 1:45 p.m. PST and 5:46 p.m. PST: For some Mac OS X 10.6 Snow Leopard users, Gears doesn't work, hampering use of Offline Gmail.

Google initially said Tuesday that Gears doesn't work on Snow Leopard but later amended its statement, narrowing the problem to Apple's Safari browser.

"It turns out that Gears does work on Firefox for Snow Leopard and Leopard, though it still doesn't work on Safari," Google spokesperson Victoria Katsarou said. "There was a bug that was preventing Gears from downloading on Snow Leopard, but we're fixing it and at we'll be updating our Help Center article and Download page to reflect the change."

She declined to comment on when or even if an HTML5-based version of offline Gmail might arrive.

Offline Gmail didn't even work Tuesday in the new Chrome for Mac beta version. Gears is built into Chrome, but trying to enable offline Gmail with the browser yields a "browser not supported" error message.

For the rest of you, here are Google's instructions for getting set up with offline Gmail:

1. Click the "Settings" link in the top-right corner of Gmail.

2. Click the "Offline" tab.

3. Select "Enable Offline Mail for this computer."

4. Click "Save Changes" and follow the directions from there.

Monday morning, former TechCrunch partner Fusion Garage revealed details of its plans to release its Linux-based Web browsing tablet.

Known as the CrunchPad until TechCrunch founder Michael Arrington announced on his blog that Fusion Garage had removed his company from involvement the product, it was expected to be a touch-sensitive slate computer designed for browsing the Web. It was said to have no local storage aside from what was necessary to load and run the operating system. Arrington said he was hoping to bring the product to market for under $300, but did not expect it to be a big seller.

More recently, Arrington said litigation over the breakup was imminent.

Fusion Garage has been quiet about the public but one-sided airing of the two companies' disagreements until now.

In a Web videocast Monday, Fusion Garage CEO Chandra Rathakrishnan laid out his position on the drama, and revealed plans for the release of the product.

Tasty hardware specs

Rathakrishnan said the product, now named JooJoo, will soon be available for preorder for $499, well above the sub-$300 price point that Arrington hoped to deliver the product at. Shipping will start in 8 to 10 weeks.

Regarding the high price, the Fusion Garage CEO said that "nothing worthwhile can be delivered," at the hoped-for $300 price, and pointed to other products, like the Kindle DX and the iPhone, as examples of products with lesser screens that cost more.

The JooJoo will have a 12.1-inch capacitive touch screen, limited local storage for caching data, Wi-Fi but no cellular connectivity, and will boot directly to the browser in nine seconds, Rathakrishnan said. It is said to have five hours of battery life and weigh under 2.5 pounds. It has no physical buttons other than the power switch, and looks quite sleek in black, its only color option. The operating system/browser hybrid, written by Fusion Garage, will be capable of displaying full HD video on the device.

Regarding the potentially competitive Chrome OS, Rathakrishnan said that while it's a "similar vision," the Netbook form factor (which Chrome OS is written for) is not ideal for the use cases he envisions. Furthermore, Chrome OS is a year from release; JooJoo is entering production now.

Ugly legal posturing

Rathakrishnan opened his press conference by saying that "The death of the product has been greatly exaggerated." He then spent several minutes refuting Arrington's claims of ownership over the intellectual property inside the JooJoo device.

"I'm not the person I've been portrayed as in the blogosphere," Rathakrishnan said in defense of himself, before attacking Arrington. "Anybody can write blog posts," he said. "There are dreamers and there are doers."

Rathakrishnan said that he and Arrington had agreed on a vision of the product early on, and that Arrington had said he would line up funding and other help to bring the product to market. The ultimate goal was for a TechCrunch subsidiary company, CrunchPad, to buy Fusion Garage. But the TechCrunch money never materialized, and Fusion Garage raised money on its own, built the product independently, and lined up manufacturing for the product.

It's "ludicrous," Rathakrishnan said, to think that TechCrunch owns any intellectual property in the JooJoo. Rathakrishnan said that there were never any legal agreements signed to that effect.

"We took all the risk and did all the work. Michael Arrington sat back," Rathakrishnan said.

Opinions differ on Rathakrishnan's position, of course. Arrington has laid out a compelling case of betrayal in his blog posts on the topic, and TechCrunch readers have, to date, supported Arrington's position that he is the injured party.

Ultimately, it's highly unlikely that the rancor between TechCrunch and Fusion Garage will make much difference in the success or failure of the product. Neither TechCrunch nor JooJoo is a popular consumer brand, and the market for this category of product is untested. What is more likely to trouble the product's success is that it's priced out of reach of most consumers, and functionally doesn't offer much more than a $300 Netbook, although arguably it performs many of the same functions with a lot more style.

We're getting a hands-on look at the JooJoo later Monday and will report on that in a separate post.