IP: WAP's up!

[Dave Farber <farber@cis.upenn.edu>]

>From: Janos_Gereben <janos@netcom.com>
>Subject: WAP's up!
>To: farber@central.cis.upenn.edu
>Date: Fri, 28 Jul 2000 18:13:02 -0700 (PDT)
>X-Mailer: ELM [version 2.5 PL3]
>
>New shots fired in heated WAP debate
>Janos Gereben - the451.com
>
>San Francisco - Responding to reports published here Thursday about
>problems confronting development and marketing of the wireless application
>protocol, a prominent scientist told the451 that "within a few years, WAP
>may fade away like a bad dream."
>
>Petri Mahonen, director of the Wireless Internet Laboratory of the Finnish
>National Institute of Technology, primarily holds "marketing people"
>responsible. "They have gone ballistic, selling WAP as 'wireless Internet'
>or 'Web in your pocket,' which is creating a backlash effecting serious
>academic and industry research and development," he says.
>
>Mahonen, a research professor with the University of Oulu, is highly
>critical of WAP implementation in his own country (Finland is the leading
>user of wireless phones) and elsewhere, painting a picture of a confusing
>situation in which defective equipment and server products are shipped, WAP
>phones are incompatible, service is painfully slow and over 30% of WAP
>connections fail  while users are charged for airtime anyway.
>
>Alan Reiter, the subject of yesterday's interview that prompted Mahonen's
>statement, agrees with some of the criticism, but puts a different spin on
>the situation: "There's no doubt that manufacturers have had a difficult
>time developing bulletproof WAP products. They admit it, and they also
>admit problems with incompatibilities. Manufacturers and carriers all
>understand that incompatible equipment will destroy WAP's chances. But the
>early days of all technologies have problems. When cellular was first
>introduced in the US in the early 1980s, only a few manufacturers were able
>to develop what was then an extremely sophisticated handset."
>
>Mahonen says that development of global GPRS (general packet radio service)
>will make connections more reliable but won't improve connection speed
>substantially. "You might be getting 10-40 kilobits per second, but that's
>it," he says. There is also the question of price, with the possibility
>that GPRS will be costly, at least initially. (Mahonen sees an alternative
>for the technologically advanced user to run VoIP over GPRS.)
>
>Besides concerns about WAP security (shared by almost everybody in the
>field), Mahonen's basic objection to the protocol is that "its philosophy
>is 'walled garden,' while its champions pay lip service to IP." The wall,
>he says, has to do with the fact that most WAP operators restrict access to
>their own WAP gateways and those of their partners, "so forget free surfing
>like on the Internet  with WAP, you are in the typical closed telco world."
>WAP is not an open standard, he charges, but a proprietary system that is
>not TCP/IP-compatible, and "still unreliable" at that.
>
>Scott Goldman, CEO of the industry association WAP Forum, repeats Reiter's
>point about the technology presenting an "evolving situation," and says
>that Mahonen is "just plain wrong" in all his charges. But he accepts the
>'walled garden' statement, pointing to the America Online paradigm. Just as
>AOL has achieved great success by providing a package of propriety
>capabilities for people who are not knowledgeable about technology, Goldman
>says, WAP offers a packaged, user-friendly service. In the future, however,
>"carriers can distinguish themselves" by providing alternatives, he adds.
>
>Regarding compatibility, WAP Forum claims that WAP will work with GSM-900,
>GSM-1800, GSM-1900, CDMA IS-95, TDMA IS-136, and 3G systems  IMT-2000,
>UMTS, W-CDMA and Wideband IS-95.
>
>For WAP security, Baltimore Technologies, Solomon Technology, and in
>Mahonen's own country, F-Secure and SSH among others, are reportedly close
>to releasing full-strength 128-bit encryption and authentication for WAP
>servers, with content security programs to follow. Goldman says Mahonen is
>"not in the real world" because high-profile companies are already using
>WAP for secure transactions.
>
>As far as writing new code for WAP, Mahonen says developers "have already
>gone and redefined almost every stack available in the UDP/TCP/IP world, so
>forget IP compatibility and philosophy." He repeats the charge that WAP is
>just an SMS (systems management server)-with-menus, and meant to be a telco
>system. "Let's not fool ourselves that it is providing wireless Internet in
>the technical sense." Goldman contradicts that: "SMS has nowhere the
>capabilities and resources you get from WAP. There is already a tremendous
>amount of content out there. Cellmania, for example, has come up with a
>list of 5,000 WAP sites."
>
>Even if WAP is improved and becomes successful, Mahonen sees it as
>"transitional technology," especially as WLANs and 3G become more
>ubiquitous. With that technology, speeds are expected in the 40-100Kbps
>range, with a theoretical maximum of 384Kbps, with more standard,
>IP-compliant connections to new terminals than WAP can provide, Mahonen
>says.
>
>
>
>
>--
>--------------------
>janos@netcom.com, SF
>Attachments to janos.gereben@the451.com