Does Mozilla Have a Speed Problem?
Mar 15, 2011 5:00 AM PT
The Firefox Web browser is about to go where it has never gone before. Firefox makers are changing their developmental strategy. This new approach will warp Mozilla.org's flagship browser through several release generations in less time than it took to advance from Firefox 3.0 to the not-yet-released Firefox 4.0
By comparison, Google has shipped a new version of its Google Chrome Web browser about every three months. The other major browser developers -- Microsoft and Mozilla.org -- have been locked into a new version release cycle of once every year or two.
The makers of Firefox, however, are posed to make up for lost time. But could their new marketing strategy be too much too late?
The Firefox Wiki recently revealed that Mozilla.org plans to release Firefox 4 in 2011, along with Firefox 5 and 6, all this year.
Firefox director Mike Beltzner pens The Roadmap, a blog about the company's Firefox product. In a Feb. 7 introduction, he revealed the new timetable for upcoming Firefox release cycles. He charged his team with the need to dig in to remain competitive with its popular open source Web browser.
Beltzner's team has been excising the gremlins from Firefox 4 since the beginning of 2010. If you believe the rumor mill, users will have it by next month. But if the Mozilla folks deliver the next two full generations of Firefox before year's end, as Beltzner promises, will users be shortchanged with buggy new code? If the developer cycle is rammed into hyperdrive to meet these compelling delivery dates, will the new version releases be little more than a numbers game that swaps out new features delayed now for later release?
"Firefox is seeing a lot of pressure to keep up with Google Chrome. Microsoft has a lot of money to throw into IE9. In order to remain relevant, Firefox has to do something. Firefox had to find ways to stop the other two," Scott Testa, marketing consultant and professor of business administration at Cabrini College in Philadelphia, told LinuxInsider.
What's the Price of Perfection?
Clearly, slippage in its market share and rising competition is driving this apparent change in Firefox's new version release cycle. But Mozilla.org officials did not respond to a request to discuss its release cycle issue.
Once the new kid on the block, Firefox versions 1.0 and 2.0 drew increasingly larger numbers of users away from their default Web browsers -- mostly Microsoft Internet Explorer. Team Firefox effectively heated up the long-cooled browser wars.
"We succeeded in re-energizing the browser market, creating competition and innovation which benefits Web application developers and users alike," Beltzner wrote in an introduction to The Roadmap. "This newly competitive market has presented challenges for the continued success of Firefox, and in 2011 we must ensure that we can deliver a product that is compelling to users in order to continue to be able to demonstrate our vision for the Web. This roadmap outlines our planned strategy and direction for Firefox in 2011."
When you consider all these variables, it makes sense that Mozilla concluded it had to do something more radical, noted Testa. They have to put out more releases to stay relevant. Google is beating Firefox in application speed. Also, it is exceeding Firefox in the use of extensions.
However Firefox developers look at their marketing situation, they stand a chance of getting burned. From the company's viewpoint, suddenly going from a cycle of infrequent releases with heavy beta testing to a rapid-fire schedule of new x.0 releases can be risky. Aside from the technical dangers, releasing a new version of Firefox every few months might not serve to re-enliven the browser in the eyes of users the way Firefox officials hope it will.
A lot depends on generating continued acceptance for the Firefox Web browser, Testa agreed. No one thought that Chrome would gain such a big user share so quickly.
"Whether it's real or imagined, the perception is that Google is paying a lot of attention to the browser market. Microsoft doesn't, and Mozilla is in the gray area," he said.
The bottom line is that Beltzner and his development team are banking on user loyalty. Changing release strategy is fraught with benefits and drawbacks. Users have to see the frequent releases as a sign that Firefox is quickly developing and is always on the cutting edge, but they cannot view the race from version 4.0 to 6.0 in less than 12 months as a series of annoyances with not-ready-for-prime-time releases.
Software development usually follows one of two methodologies. In a waterfall approach, developers conduct a design review to cover the entire application at the end of the design phase. They use this as a checkpoint before entering implementation.
In an agile approach, developers conduct a design review as a part of every sprint that had design impact. Developers do larger and more intensive reviews in early sprints and taper off (replaced by code reviews and other activities) as the sprints spend less time on design and more time on implementation.
"The idea of switching to a breakneck release pace is essentially moving to an Agile development process. As with most things, there are both advantages and disadvantages to using an Agile methodology versus another approach, such as Waterfall," Jason Taylor, CTO of Security Innovation, told LinuxInsider.
Both the Waterfall and the Agile approaches are independent of security. However, the Waterfall method is falling out of favor, according to Taylor.
"The heavy reliance on Web applications is more conducive to the Agile method. It lets the Firefox Web browser developers concentrate on high priority updates and then push it out fairly quickly. This is preferable to coming out with a lot of big-bang releases. I really identify that with the Agile approach," said Taylor.
He does not know for a fact that Firefox's developers are using Agile development, but he would not be surprised if they were since it fits what he sees happening, Taylor explained.
"I used to work on the team for IE 4, 5 and 6. I know we didn't follow the Agile approach then. I'm pretty sure it hasn't changed. From a user's perspective, the Agile approach lets them get needed improvements in small chunks so they don't have to wait for major releases. It holds them to the product. They can see improvements over time, and it makes them less likely to look for another product," he said.
The Mozilla team was taking too long between releases after building up so much momentum, added Testa. Statistics shows that more and more people are browsing via smartphones. That shift in user focus could be a new frontier for the Firefox browser, he suggested.
"It's no longer a browser war at the desktop. Now developers are fighting for the smartphone browser market," said Testa.
Without a series of new releases, Firefox could fall to the wayside, much like the Opera browser. In fact, Opera has almost dropped off the map except for mobile products. And Opera's developers do not cycle new releases so fast, noted Rob Enderle of the Enderle Group.
"The marketing strategy is changed. It is no longer like the old days when you kept coming out with new releases to keep potential new users focused on your product. It's no longer like that," Enderle told LinuxInsider. "So you need that balance between keeping your product fresh and letting your users get comfortable with it. Anything more than once a year is problematic."
No Choice But
Mozilla developers have run out of options. They must play catch up fast or risk losing Firefox's ranking as one of the top three Web browsers, according to Chris McIlvoy, solutions architect for Siteworx. It can't afford to be the next Opera by dragging its heels, he said.
Firefox cannot do multi-process like Chrome, and IE10 is now getting some pretty favorable test results, he said.
"So Mozilla has to stay competitive through technology updates, even if it antagonizes some users," he told LinuxInsider.
Against All Odds
Mozilla is spread thin, and its ability to market improvements limited, warned Enderle. Doing a lot of release cycles tends to dilute the user excitement further and puts the product in perpetual beta mode. In turn, that could create increased security and reliability problems.
That reputation could cause businesses to flee Mozilla's browser products. So it is best to keep releases infrequent and patch frequently. That way they can better assure quality, he said.
Once a year is likely more than fast enough to stay current and still provide a period where the product can maintain stability, he added.
The success or failure of Mozilla.org's planned switch to multiple, rapid new version releases within this year could ultimately hinge on the end users, McIlvoy guessed. Rapid release cycles may be the future anyway.
"When you look at your release strategy, you first have to look at your audience. How does their intolerance for updating compare to their demand for technically superior performance or functionality? When you look at Firefox today, its browser is going to have higher demand for cutting-edge features, and less apprehension about new release bugs," he said.
More casual users, however, will not have the same fearlessness when it comes to new releases. Casual users will not have the tolerance for continuous updating. Instead, they will sacrifice performance for an easier user experience, McIlvoy noted.