The Pulver Report: July 1, 1999

Heard on the Net:

People on the Move:

This section is normally dedicated to news about specific people moving to other companies in this volatile industry. However, Vocaltec (http://www.vocaltec.com/about/press/pr_Q2initiative.htm ) recently announced a restructuring that had over 40 people leaving. Furthermore, some people have left independent of the announcement, one of whom is our friend Dan Berninger. Dan left Vocaltec and has set up a systems integration company called TelecomComputers.com. We wish all of the ex-Vocaltec'ers well.

In terms of more individual departures, Monty Sharma of MTT has joined Novell to form an Application Service Provider division, and Catherine Treblink has left Lucent for Level 3. Telcordia Technologies has seen the departures of Mike Ramalho to Cisco and Chris Celiberti to Portal Software. Francois Menard is MIA after leaving Mediatrix. If anyone knows where he is, please let us know.

Companies on the Move:

Along with the wholesale migration of people we have also seen wholesale strategy shifts within the VON Industry. The last month or so has included more mergers and acquistions. Some have been based on 'filling out the product line' while others seem to be based solely on the fear of being competitively disadvantaged. The announcements show a shakeout of a sorts at time when interoperability seems to be the next big push. Many companies are improving their interfaces for the shift to next generation network elements.

Alcatel managed to traffic VPN provider Internet Devices into its portfolio for US$ 180M ( http://www.alcatel.com/press/current/1999/06_16.htm )

Cisco, still pitching the 12000 as a straight WDM integration, continues to bulk up with its acquistion of StratumOne ( http://www.cisco.com/warp/public/146/june99/47.html ) and its ATM hedge bet on Transmedia ( http://www.cisco.com/warp/public/146/june99/32.html )

Dialogic being bought by Intel for US$ 780M allows Intel to expand its strategy for server/board development. Most interesting is the fact the plan is for the Dialogic unit to become part of the enterprise business unit ( http://www.intel.com/pressroom/archive/releases/cn060199.htm )

Lucent expanded its backbone product line with Nexabit ( http://www.lucent.com/press/0699/990625.coa.html )

Saraide may be the best-positioned Application Service Provider with its acquistion of GIN as a content provider for its GSM information service (hence Ericsson's recent investment) ( http://www.saraide.com/company/newsroom/nr061499.html )

Tellabs has added NetCore to its product line to build out its NextGen network's ATM product line ( http://www.tellabs.com/new/99news/nr063099.htm ) Texas Instruments bought Telogy for US$ 435M, which is interesting since Telogy was already a solution provider for TI (http://www.ti.com/corp/docs/pressrel/1999/c99026.htm )

Unisys got the NT message and acquired Pulsepoint for US$ 100M, promising to be an interesting integration effort (http://corp2.unisys.com/AboutUnisys/PressReleases/1999/jun/06156726.html )

VoiceStream attempts to have West meet East with its acquisition of Omnipoint ( http://www.voicestream.com/about/press.htm#op )

Nuera comes of age by finally closing a deal to provide IBM with voice packet products ( http://www.nuera.com/News/pr21.htm )

Miles apart the wunderkind choose differently: In a "nonexclusive deal," Level 3 announced a five-year contract ( http://www.lucent.com/press/0699/990625.coa.html ) with an initial US$ 250M purchase of the Lucent Softswitch (hardware portion not directly mentioned), while Qwest committed to three years worth of Nortel optical network solutions ( http://www.nortelnetworks.com/corporate/news/newsreleases/1999b/6_30_999945 5_Qwest.html)

It's great to have sizzle, but eventually you have to deliver steak. The recent boom on Internet stocks has been amazing to watch. The run up and capitalization of many of the companies has made even the most skeptical bullish. Don't be suprised to learn that your next door neighbor has filed their S-1 and will be going public in the next sixty to ninety days. The members of our community who have filed or are recently out on the street during the past month include: AudioCodes, Clarent, Delta3, ITXC, Juniper Networks, Net2Phone and Phone.com.

pulver.com posts IP Telephony news daily at http://pulver.com/news. If you would like to subscribe to our weekly IP Telephony News newsletter, please visit: (http://pulver.com/news/subscribe.html )

Megaco makes progresso
The great spamming email system known as Megaco has progressed to a good compromise and will be ready to present at the IETF meeting in Oslo later this month. As explained at VON Europe last week, the new solution increases the opportunity to take advantage of multiple connections via the network in a manner that may be useful for alternative access devices such as wireless analog, CDPD / broadband and POTS lines. It will be interesting to see how quickly the protocol is incorporated by the MGCP solutions and how, if at all, the new aspects enhance the services.

VON Europe '99 sets themes for Atlanta
875 delegates attended VON Europe '99 last week in Helsinki. We choose Helsinki for VON Europe '99 because it represents a city and population which has been living with the effects of having pervasive wireless and internet connectivity.

At the conference, the theme was the convergence of Wireless with IP and the PSTN (see below). Many presentations were given on the new protocols that were blending the wireless device and the application with the lifestyle of the end user. Protocols such as WAP and Bluetooth were discussed, as well as Parlay and SIP and their relation to wireless devices.

During the Navigating the CrossRoads to Interoperability preconference workshop, there were reports on the efforts of various organizations on various protocols, trying to speed the process toward interoperability. All of the efforts are gathering momentum, though the approaches, like the protocols, vary greatly. Some groups have overlap and may eventually merge or distinguish themselves from each other. Some of them are specifically focused on operational issues such as quality. All of them should have progress to report at the next iteration of this workshop at Fall '99 VON ( http://pulver.com/von99/fall99 ) in Atlanta

Softswitch Consortium
The International Softswitch Consortium is an organization formed to promote the "global cooperation and coordination of internetworking technologies in the field of Internet-based real-time interactive communications and related applications." The intention is to collaborate with other organizations in developing and promoting "Foundation Protocol Specifications." The foundation will be an interface specification for the use of members and developers, with protocols including: RTSP, MGCP, and SIP. The by-laws are now posted. ( http://www.softswitch.org ) Other events include an interoperability event and developers' conference. The first meeting for all members is set to take place later this month.

The development of the International Softswitch Consortium [ISC] represents a possible milestone for the Internet telephony industry. The goals of the consortium are to foster third-party developers building solutions in the marketplace and to enable service providers to better match customer requirements for application-specific solutions. To help this industry as a whole requires strong tools for the developers and active dialogue.

Parlay
The Parlay Group recently announced the expansion of its membership and its Phase 2 work plan (which is to be completed by December). The group is promoting industry acceptance of the Parlay API, a specification designed to enable carriers and independent software vendors to write applications to provide services across wireless networks, IP-based networks and the PSTN. The work is intended to expand the API for wireless-based services and introduce value-added services to the IP world. The specification is aimed at attracting the community of third-party applications vendors.

Parlay's Phase 1 activities in 1998 included the design and development of the API, covering such functions as call control, messaging, and security. The results of that work were published and demonstrated in December 1998 in both the UK and US. The guaranteed call delivery demo application featured a multi-location, multi-vendor equipment configuration that showed a set of Parlay API capabilities, including service provider-controlled call delivery combined with Interactive Voice Response (IVR) and e-mail. ( http://www.parlay.org )

Wireless/IP Convergence
It is not insignificant that Third Generation (3G) wireless includes an IP layer. When IP assumes ubiquity throughout wireless, satellite, data and voice systems, then there is a uniform platform for innovation and interconnectivity. This is the grand vision for Voice on the Net.

United States based wireless providers are right now simply finding ways to bring the net to the phone with systems such as WAP. Palm VII and its strong acceptance into American culture is really just a tease for what anyone can take for granted any day in Helsinki.

But this is just the beginning. Eventually, NextGen killer apps will allow critical mass for this platform, and wireless devices will be network access nodes, not simply phones. Voice becomes an application, and in the same way that radio stations can be heard on everything from stereos to headphones, computers to clocks, telephone calls may one day be received on everything from stoves to cars. There is no reason for a dedicated voice appliance when standard IVR functionality in a device ('voice inside') can be expanded into something more broadly useful.

Furthermore, devices can gain context-sensitivity. This can work in tandem with preferences learned through phone calls made and web pages accessed. Service providers gain a competitive edge by their capacity to analyze the preferences and current status of the user and then act upon it - suggesting restaurants in the present city matching the user's tastes, permitting phone calls from web pages or automatically arranging a schedule with another person's wireless device.

This is part of the idea of 'always on' - programs such as ICQ will be expanded and given the functionality we see in products such as Ericsson and Oz.com's iPulse - allowing routing of different individuals to different forms of communication out of an array of available choices. They will even route to landlines when possible and thus integrate seamlessly with the legacy networks.

Wireless threatens the PSTN more than the Internet. Once the kinks are worked out, wireless can become the tool that allows penetration into the Third World without constructing a costly PSTN infrastructure and that will make NextGen technology worth so much more than legacy systems. The opportunities are clearly greater than the risks.

When to buy?
Vendors, friends of vendors, and others coming out of the woodwork are proudly hawking IP-based PBX's, thinking people will buy them just because they're IP-based. Truth be told, nobody cares. The underlying key to their success will be what they do differently - features and functionalities will become the market drivers compelling people to throw out their significant investment in PSTN PBX's. When these key features are available, the IP-based PBX will be the corporate change agent.

At the same time, future edge devices are going to be a fraction of the cost of the current implementations. When a chip can handle converting voice to IP, klunky gateways become outmoded. The future will be speciality devices which can be designed for PSTN, wireless, or the Internet, probably having ports for each (they may even be microrouters). From an economic standpoint, wire is cheap and IP telephony gear is cheap, so why converge? The risks associated with convergence are greater than the cost of multiple networks.

The single wire theory is dead. Here at pulver.com, the connectivity of our new office could have been built onto one wire, but we opted to have four sets instead: two for voice, one for data, and one for experimenting. Migration to new devices should not be done at the expense of a compay's day-to-day operation. Instead, test networks and labs will have to develop implementation plans and slowly integrate the new with the older legacy.

And this is just part of a larger story: people are noticing that IP is not necessarily a grand unifier. You do not have to mandate that all services being managed be IP in order to benefit from lower systems integration costs. As long as there is a gateway between legacy systems and the IP network, the costs for administering the network already become sufficiently low in terms of development, and additional value is extracted from previous purchases.

The Pulver Points on Directories (version 1.0)
In an 'always on' world the use of directories will become essential to establishing the right contact at the right time. This not only includes instant messaging but other aspects of indexing people, places and things. The Pulver Points below describe some of the trends we are seeing in directory at this moment in time. As always, comments and suggestions are welcome.

1. Instant Messaging The most visible location for directory innovation this year will be in the use of Instant Messaging. Currently AOL leads the pack with both ICQ and AOL Instant Messenger. Portal/search sites will bundle the service with other features including voice (see Yahoo! Pager). But the most innovative work will come from service providers who will bundle the service with Personal Information Managers, Portable Digital Accessories and web browsers.

2. LDAP & Schemas LDAP is quickly becoming the fastest-adopted protocol on the Internet since HTTP. It has been adopted by web sites, enterprises and service providers 97 all with different requirements. The result is that while LDAP will continue to evolve, the problem of proprietary schemas will not disappear. Companies will develop LDAP applications in the workgroup and for centralized locations; only after they begin improving management practices will they realize the extent of the problem.

3. Meta Directories Using a Meta Directory will be considered the solution for some time until it is realized that the schemas are so diverse that Meta Directories are just a central location for the headache and not the solution. Many Meta Directory solutions will attempt to establish a patch point where specific applications schemas will be offered as a "standard" in attempt to reduce the development and integration costs. Companies best positioned to lead the way in this space are the former X.500 leaders: DCL, IBM, Isocor, IC, etc.

4. DNS Address Management ICANN will be promoting the diversification of Internet DNS control by authorizing competitive registries. Domain registries are going tob explode with bundled services and new applications, prices will drop an average of 50% with some 'free lunch' services. However, few of these companies will be interested in maintaining the root system. The result: innovation will be of a business-oriented rather than technological nature for the next few years.

5. epersona & the eBill of Rights As we get more virtual the need to control your virtual entity will become more and more important. Improved epersonification tools will aim at making specific experiences (such as shopping and virtual communities) on the web more relevant to individual users. Other tools will attempt to block the use of demographic information from the end user's client. This battle will be highlighted with court decisions on virtual privacy, implied endorsements, and full disclosure requirements.

6. LAN versus email Today the average corporation maintains over twenty directories on its networks. The LAN and email are the most prevalent systems. The NDS versus Microsoft debate will continue to fade as the use of LDAP servers connected to email systems supersedes this battle. As companies build applications that integrate directory servers, the messaging model will continue to be the most relevant place to innovate. But do not expect an easy marriage, and expect that it comes from third-party innovation using API's provided by Microsoft.

7. DEN, QOS, VPN and other three-letter acronyms QoS & VPN solutions will continue to be offered based on vendor-specific technology. The use of independent DEN features will not be viable until the network management features improve. Administrative costs will be identified as the largest component of implementing a class of service-type systems. Connectivity using ERP and accounting systems will lead the way in billback and price differentiation.

8. Caching Tools and Synchronization Edge management will need to develop an interface between support systems and network interfaces. All signs point to this being the most important work, and nowhere does it seem to be ready now.

9. Search & Seizure The web has revealed that content is not king, but a driver to communicator and community. This runs in the face of the portal content alliances that have been formed by AT&T and others. Service providers say they get it, and yet we see intertwining of content and sales when the stronger model is community that creates content.

The pulver.com Open Test Network: Update
Now that we have completed the move to the second floor of 115 Broadhollow Rd., Melville, NY 11747, the Open Test Network is open for operation. The racks are set up, and the facilities include two PRI's and a T1 to the Internet. At VON Europe, with little more than a week to prepare, we developed two test demos.

Thanks to Henning Schulzrinne and John Lennox of Columbia University we have a SIP server, ( http://sip.pulver.com ) and we demonstrated its interoperability with generously-provided SIP APA's from Mediatrix. The demo used the Internet to connect Helsinki, Finland, Melville, NY and the Mediatrix offices in Quebec to complete basic phone calls.

The second demonstration was with the thin clients associated with such products as Buddyphone and Qtalk. Thanks to Andrea Electronics's headsets, we showed the value of echo cancellation and silence suppression.

At the show, RADCom loaded their diagnostic software so we could display information about the quality of the service.

At Fall VON in Atlanta we expect that the demos will be expanded considerably, and we look forward to other participants. If you wish to participate, we remind you that the objective of the Open Test Network is to be a presence on the Internet available to demonstrate to customers and alliance partners and to work collaboratively on interoperability and other concerns. It is not a location for terminating traffic. Our primary function is to act as facilities coordinator. We can provide the space, hands and eyes to place devices in the network. We will maintain a run log for the devices.

We are looking for a switch to reduce the cost of PRIs. Donations would be welcome. ;<) For more information please visit ( http://pulver.com/testnetwork )

We've Moved
pulver.com has moved. Our new address is: 115 Broadhollow Road, Suite 225, Melville, NY 11747. Telephone: +1.516.547.0800, Fax: +1.516.396.7870