BREW This - With the proliferation of Java devices in the worldwide wireless data services market, what are BREW's chances of ge
BREW This - With the proliferation of Java devices in the worldwide wireless data services market, what are BREW's chances of ge
By: Robert McGarvey
Jan. 1, 2000 12:00 AM
Remember back when there was a real and fun! battle for supremacy between Microsoft and Apple? Guess what: just such a conflict is shaping up for supremacy on the wireless handset where at least two major players are girding up for a fight to own this turf. Of course there's Sun with their lightweight, mobile Java (J2ME), but there's also QUALCOMM with BREW. Who's ahead? Read on for eyewitness reportage from this battlefield.
Just when you thought the battle for the handset was over and it was time to sip a ceremonial cup of Java in honor of Sun's plans to dress up handhelds with Java 2 Micro Edition (J2ME), here comes a challenger with an exceptional pedigree, real code that already runs on handset chipsets, and an immense ambition to sit in a perch where it will control what applications flow from developers to carriers to handsets to users.
The challenger: BREW. The maker: QUALCOMM.
Launch date: you're looking at it. Around the time you read this, Verizon will be unveiling BREW apps as well as phones across the U.S. in time, the company hopes, to tap into the Christmas shopping season. A Korean launch is slated for November 2001. A Japanese launch is on the schedule for late winter 2002. More U.S. launches (Leap, Alltel) will follow in 2002 and also in planning stages are BREW deployments in Brazil and New Zealand.
BREW Apps Galore
Compare that smorgasbord to what's probably on your handset now a wimpy scheduler, anemic e-mail, a few bland games that are boring before the first play is completed, maybe a calculator, and a few other unexciting apps that users, so far, have succeeded in ignoring.
BREW plans to be different at the very start by offering both practical and entertaining apps to carriers, which, in turn, make these apps available to subscribers. QUALCOMM hopes BREW will become a popular environment in which users will rack up many minutes playing games and sending messages. In fact, QUALCOMM's revenue-sharing model, splitting proceeds from subscriber's application purchases, attracts the company's key target markets: wireless carriers and application developers. QUALCOMM's BREW additionally gives developers from smaller foreign outposts to established "big dogs" a level playing field as they seek to sell carriers on their latest creation. Carriers win too, with markets beyond just consumers to enterprises, since BREW sets the stage for easily making a business app available to, say, all BP or Disney or Schwab employees carrying a company cell phone. That same app could be unavailable to anybody else because exclusivity for carriers as well as customer groups will definitely be for sale since, from the get-go, QUALCOMM has envisioned BREW as a wireless cash register that will ring in revenues for carriers, app developers, handset makers and, of course, QUALCOMM, too.
The Underdog Barks
What's more, while Java may have succeeded with current deployment, QUALCOMM is winning wide telecom industry support not surprising because QUALCOMM is one of them, not an outsider from Silicon Valley. QUALCOMM is a powerful known entity in the wireless business because they provide the majority of chipsets used by CDMA phones. While GSM may currently be able to claim more subscribers, the reality is that CDMA still rules in much of Asia and even in the U.S., where two of the leading carriers Verizon and Sprint are CDMA-based. And CDMA is the basis for both flavors of the coming 3G standard CDMA2000 and W-CDMA.
CDMA isn't going away soon. "Our estimate is that 80-million CDMA handsets will be sold worldwide in 2002," says Peggy Johnson, president of QUALCOMM Internet Services, the division that has masterminded BREW.
A big BREW selling-point is that it will be on most, maybe all, of those CDMA phones because, says Johnson, it runs every bit as well on a cheap handset as it does on the premium units. On the Java side, handsets only require as much memory as a WAP 1.2 browser to run Java apps, and the lowest Nextel Java phone in the U.S. is now selling for only $50. "BREW can slip right on low-end phones," says Johnson. "In fact, we developed it to run on low-end phones and later saw it had value on higher-end phones, too. It's not at all CPU intensive and doesn't need much memory."
"The developer has this choice," says Johnson. "He or she can write an app that will run on all BREW handsets, or a Java app that will perhaps run on the top 20% of handsets with more memory."
That, suggests Johnson, is an easy choice and the bottom line, she says, is that BREW offers a credible value proposition to everybody who wants more functionality on cell phones. To handset makers, it's a freebie and QUALCOMM will work even with OEMs "to help them integrate BREW correctly," says Gina Lombardi, senior vice president of marketing for QUALCOMM Internet Services.
A Developer's Bonanza?
How? According to Johnson, carriers can pick BREW-certified apps, hammer out deals with developers, and then all monies flow back to QUALCOMM from the carriers, with "80% of revenues going to developers," says Johnson. (Note: Carriers will do all the collecting because they already have a financial relationship with the customer and adding a new fee for BREW apps to the bill is simple, just a matter of a few keystrokes.)
One hitch: BREW apps are written in C++, an old code that is lean "apps need to be lightweight to fit into a small space," says Johnson but C++ also is a stubborn code that many developers long ago abandoned in C++ favor of the easier Java. Will there be enough developers willing to write in C++? "You bet", says Johnson, who is plainly betting that the lure of easily collected authorship fees will bring around the development community.
How will these apps be priced? According to Johnson, some will be free (subsidized by the carrier), some will involve a one-time fee, others will involve a recurring monthly fee, perhaps $1 or so. Most handsets today in the U.S. will accommodate only a few apps, but this still might turn into a bonanza for developers because, says Johnson,"with BREW, a user might own 100 apps, with perhaps four on his handset and the other 96 on a server, available for download." Tired of playing that baseball game? No sweat jettison it and download a football game. "BREW lets users customize their experience," Johnson adds.
The remaining 20% of revenues is split between the carriers and QUALCOMM Johnson won't reveal specifics, other than to indicate that the exact percentages vary from deal to deal and, for this reason, carriers, too, are warming up to BREW. "BREW is a new value-add carriers can offer their customers," says Lombardi, and it delivers on two levels for them. "It may help carriers reduce churn," she says, and that's because if a user develops a fondness for say, a Verizon-exclusive app, he'll be unlikely to jump to a competitive carrier to save a few dollars.
Additionally, "BREW will be a new source of revenue," says Lombardi, and for carriers who have wrestled with the disappointment of WAP (at least in terms of revenue generation), increasingly treacherous low-end pricing of buckets of minutes, and mounting build-out costs for next-generation wireless infrastructure, any promise of new money is welcome.
"Carriers have been very receptive to BREW," says Johnson. "From the beginning, we identified carriers as our customers. We went into this very carrier-centric and, frankly, that's a key differentiator with BREW."
Broadening the Battlefield
Lombardi echoes that: "BREW is not uniquely CDMA. It can be used across technologies and we support that."
QUALCOMM is just as adamant that BREW will hold wide appeal for handset makers and, while they decline to name names, they say that at least three different makers will be represented at the Verizon launch and more will be joining the party soon. "BREW will give users a reason to buy new more powerful phones, perhaps ones with color screens," says Johnson. Up until now, users generally kept a phone until they lost it, broke it, or sometimes a carrier offered a sweet (and low-cost, even free) handset upgrade, meaning that handset makers have been stymied in creating the kind of brisk replacement market that has long fattened the coffers of computer makers.
And this use of BREW to nudge users to upgrade handsets will be true from the launch on. "We have no current plans to make legacy handsets BREW-compliant," says Johnson.
Add up the QUALCOMM analysis and the bottom line is that many groups stand to benefit enormously from BREW: carriers (who reap more minutes, more revenue, and more customer loyalty); app developers (who win more revenue); handset makers (more sales of high-end phones); end users (more personalized, useful handset content); and of course QUALCOMM, which if BREW flies positions themselves as an intermediary that grabs a small but meaningful slice of revenue whenever an end user puts a BREW app into service.
That's strongly worded but McAteer's point is that longtime QUALCOMM competitors such as Nokia, Ericsson, and Motorola won't likely be persuaded to support a QUALCOMM initiative. McAteer is dubious that former handset competitors such as Nokia, Ericsson, and Motorola will support the QUALCOMM initiative (though in fact, Sony, Ericsson, and Motorola Korea have signed BREW MOUs). "I just don't see it,"says McAteer, and while QUALCOMM has won support from some Asian handset makers notably Kyocera, which owns rights to the former QUALCOMM handset line the big players will ignore BREW, according to McAteer.
Equally skeptical is research from Forrester, in a report titled, "QUALCOMM's BREW Not As Potent As Java." The document, principally authored by analyst Bob Zurek, agrees that "handset vendors Nokia and Ericsson will not see BREW anytime soon." That is, BREW's primary focus will be the CDMA universe.
A key point made by Zurek is that developers may not welcome the idea of QUALCOMM's 20% cut (for themselves and the vendor) of fees earned by apps. "Wireless vendors already taken to task by their investors to get to profitability will have a difficult time justifying a 20% cut of their product revenue." He adds, "Skip BREW because Java will do QUALCOMM has not offered a compelling motivation for adoption."
Still more scathing is the Giga Information Group report, authored by analyst Carl Zetie. He says: "The biggest limitation in considering BREW as an application development environment is that it offers portability only across QUALCOMM CDMA handsets. This is a degree of portability that is without value to the average application developer, who does not (and should not) care whose ASIC chip is running in a phone!"
Zetie goes on: "The BREW platform also introduces a great many dependencies on QUALCOMM itself for the application developer, which will make it unappealing to the enterprise. QUALCOMM provides the software development kit (SDK), runs the certification process, and generally acts as a gatekeeper for BREW applications. Few enterprises will want to put QUALCOMM in such a position of control."
How? According to ResearchPortal.com, BREW has a substantial advantage. It was designed from the ground up to work on voice-centric devices. "Java was not designed specifically for wireless, or for that matter, mobile technologies. Worse, "Java was not designed to be stretched to the extent that is required for certain classes of voice-centric SFF [small form factor] devices," meaning that it's very difficult to tweak Java code so that it can run, robustly and with stability, on a low-end handset.
BREW, concludes ResearchPortal.com, is built with an "inherently mobile design" and, therefore, is more compelling "for next-generation voice-centric SFF phones."
Another voice that at least tilts toward a pro-BREW position is that of analyst David Hayden, CEO of research firm MobileWeek: "BREW is a much better approach for monetizing these applications on mobile phones," and that's because in the J2ME world, where there is no central clearinghouse, each carrier or handset maker or both will have to certify apps for their users, and neither is set up to make this chore a high priority.
"There will be 10 times more Java apps," adds Hayden, "but at the end of the day BREW has the advantage because it makes monetizing easy" and that, suggests Hayden, will attract developers of in-demand, heavily used apps.
Another BREW advantage, according to Hayden: "I guarantee you that QUALCOMM has a better relationship with CDMA handset makers than does Sun," and, importantly, while QUALCOMM may have no particularly strong relationship with non-CDMA manufacturers and carriers, neither does Sun.
"The carriers," adds Hayden, "like the idea of BREW" and, so far, most are silent about Java.
Will BREW dominate in Hayden's view? That, he says, is too early to call and two huge questions must first be answered: "Will CDMA spread quickly enough?" and crucially, "Will end users actually want to pay for these services?" Nobody knows.
Certainly in Japan, for instance, users happily pay for NTT DoCoMo data products (games, calendars, etc.), but Japan is not necessarily a valid test bed for what will fly in the U.S. "We still have yet to validate the notion that U.S. consumers will pay," says Hayden and, in that regard, BREW may stand as the wireless industry's first broad test of consumer and enterprise willingness to pay extra for wireless data products.
Just that may in fact be the tipping point when it comes to identifying the winner in providing apps to wireless handset users. First there has to be clear proof that there's a sizable user base willing to part with cash to access these apps. Then it will be time to start picking winners and losers but, until that market is validated, it's way too early to say that BREW or J2ME, for that matter, is the leader.
Monetizing Wireless Data
Of course QUALCOMM has their own motives they want to stimulate sales of CDMA chipsets and, in the process, are eyeing what could become a gusher of cash that rains down on the middleman in promoting wireless apps. "QUALCOMM's effort resembles the record industry tariff extracted from musicians," says Forrester. Another plus for QUALCOMM is that by virtue of their middleman role, "QUALCOMM will collect valuable information on which applications are succeeding," says Forrester, adding that QUALCOMM, in turn, will use this info to make savvy venture capital investments in wireless app developers.
Some analysts take these facts as evidence that QUALCOMM is fast on their way to becoming the wireless world's bullyboy. But just as credible at this stage is the notion that QUALCOMM is to be applauded for taking the first real stabs at making wireless data a reality for the mass market.
In the meantime, QUALCOMM shrugs off the negative analytical commentary by pointing out that, as far as the broader audience is concerned, BREW's problem isn't skepticism but simple lack of awareness. "There's not enough buzz around BREW that's been a problem for us," says Johnson. Not enough analysts or developers or consumers or enterprise IT managers, for that matter, concedes Johnson know the BREW story in detail. When they learn more about BREW, Johnson says, they will be swayed by the benefits and, she adds, with a major advertising-marketing campaign planned by Verizon to kick-off its BREW initiative, awareness is about to be raised big time.
Will that thrust BREW into an undeniable number-one slot? While clearly wanting greater awareness for BREW, Johnson won't throw down that gauntlet. Short-term at least, QUALCOMM's goals are smaller. "We are not competitive with Java. We are not competitive with anyone, really," she says. BREW's goal, she reiterates, is to get apps on as many handsets as possible. And in the process, she says, "we know that we are new and we have to prove ourselves, to prove that BREW will deliver as promised. That is what we will be achieving in the near-term."
J2ME: The Sun Also Rises
If there's a 900-pound gorilla in the wireless applications space it's J2ME, the lightweight edition of Java designed for a mobile environment. Here, a for-instance of J2ME's marketplace stature: in 2002, Nokia alone says it will ship 50-million Java-enabled handsets. Motorola is already shipping Java-enabled phones, and says all their phones will be Java-capable by the end of this year. Nothing else comes close to matching these numbers not BREW, not anything. "The number of Java-enabled phones will keep growing," says Eric Chu, a Sun group marketing manager who heads the mobile device push.
Sun has an undeniably large head start in the wireless world: by late summer, not a single BREW device had shipped, but already "over 8-million Java handsets are in use," says Chu. A big presence in the U.S. is Nextel, which last spring began shipping Java-enabled handsets. Bottom-line: J2ME is definitely working. "That's a fact," says Chu.
Stumbling into the Lead
So how has Sun pulled off becoming the platform to beat in the wireless world? "By being humble," says Chu. "We knew we were coming into this new to wireless data services, so we set out to pull the best minds together to shape our efforts. We went to Nokia, Motorola, Symbian, Ericsson, J-Phone, NEC, Siemens, RIM, Sony, and many other key players to solicit their support and to ask them to lend their expertise in putting Java to work on handheld devices. That's a big factor in how we got to where we are."
Another factor: "Java is an industry standard, it's not the product of one company," says Chu, and this widespread support, he suggests, is why broader Java adoption on cell-phone handsets is something of an inevitability.
Plus, he indicates, Java has definite advantages. For developers, "it's much easier to write Java than C++," says Chu, who proves his claim by pointing to the literally tens of thousands of developers working on Java apps for small form factor devices. No competing technology has a development community anywhere near the size of the Java developer group.
For carriers, too, a plus is that "Java authenticates apps line by line; nothing else offers the same kind of security," says Chu, who indicates that this means users can download Java apps on the fly to their handsets and do so without worries that the applications may harbor viruses, worms, or worse. "C++ and similar languages don't offer that level of security and this is important in the wireless world. Java is the only application that's safe."
Industry analyst Seamus McAteer, with Jupiter Media Metrix, agrees: "Java allows protection for other key functions of the handset."
But will there be carrier support? So far, carriers in the U.S. that have announced support for Java include Nextel and Sprint PCS. The big boys Verizon, AT&T, and Cingular either have made bets elsewhere (Verizon on BREW, for instance) or they are staying on the sidelines until it's clear that the wireless infrastructure has gotten muscular and slick enough to handle substantial volumes of data. Sun is very much aware of this lag: "We are working with carriers," says Chu. "We are trying to help the industry pull together the pieces to really enable wireless data services."
Growing the Wireless Data Market
Users, he adds, will turn on to J2ME as soon as they get Java-powered handsets and begin to see the benefits. Already, for instance, Nextel customers can get corporate e-mail routed to their phones (Java software interfaces with enterprise servers), cache Web-based calendars on their phones (so appointments can still be checked even if there's no network connection), and play dozens of spiffy games. Once users turn on to this power, they'll want it, particularly because of the cornucopia of Java apps that are now customized to run on handsets and they will likely be joined by many more apps in coming months as developers rush to enter this market.
That's Sun's vision, and when you ask developers, most agree with this J2ME forecast: "There are lots of Java developers out there. Sun has been very aggressive about recruiting developers," says Pascal Chesnais, CTO at MessageMachines, a Boston-based wireless middleware developer.
Adds Scooter Willis, a wireless engineer with Fourbit, a Fort Lauderdale infrastructure company: "J2ME is everywhere in the development community. It is winning lots of attention."
Nobody else definitely not QUALCOMM with BREW has achieved this top-of-mind status with developers. This said, however, Sun insists they want no battle with QUALCOMM or BREW: "We see the platforms as complementary. They can run on the same handset," says Chu. "They can sit side by side."
He adds: "That QUALCOMM has decided to enter this market helps to validate it. We don't believe J2ME has any competitors. Nobody has the breadth of offerings we do."
And it's plain that Sun has no intention of relinquishing that perch.
Reader Feedback: Page 1 of 1
Latest Cloud Developer Stories
Subscribe to the World's Most Powerful Newsletters
Subscribe to Our Rss Feeds & Get Your SYS-CON News Live!
SYS-CON Featured Whitepapers
Most Read This Week