THE PULVER REPORT: The September 16, 1999 Issue
In this Issue:
Heard on the Net:
- People on the Move
The Month of September '99 has seen the continued shifting of some of the people
who have played a role in the growth of the international VON Industry. Amongst
those on the move this month:
- Tim Howes (father of LDAP) left AOL and is rumored to be working with Marc
Andreessen on a hush-hush new venture in the Valley.
- Eric Zimmerer left Level 3 and is now working at Ipverse - Rick Pierson left
Mockingbird and is now also working at Ipverse
- Eckhart Gouras has left POPTel and is now at Pacific Telekey Network
- David Sokolic has left Vocaltec
- Companies on the Move
I may have been right that 1999 was the "Year of the Application" as evidenced
by the growth of applications such as the Softswitch, but it is also the year
that VON met Wall Street.
During the past couple of months a number of VON Industry players have filed their S1s and will be coming out over the next weeks and months.
Notes and Thoughts from the premiere Instant Messaging Summit
On September 8th, I hosted the first major gathering of the Instant Messaging
[IM] industry in our new offices in Melville, NY.
While preparing for the meeting, I started to think about the importance of Instant Messaging and in particular I started thinking about Presence Management as a key enabling technology for service providers to increase call completion. As IM technologies evolve, look for IM applications to become the "killer" application(s) which will justify any service providers investment in IP Communication technologies.
At the meeting I was asked by some of the delegates not to mention the names of the companies that attended. What can be said is that it was a good mix of the current instant messaging industry even though, because of scheduling conflicts, AOL, Novell and Yahoo! were not there. In total, twenty-seven companies were represented.
It was clear that one of the common goals of the attending delegates was to make sure that the group stayed focused on the standards process. A key to the industry's future success will be the ability to achieve interoperability ( and operability ) within the IM space.
One thing which I found interesting was that some of the attending companies had already achieved interoperability utilizing existing IETF standards. As I discovered from the first wave of feedback emails I received after announcing the Summit, the protocol that people use really is not the issue. Rather, at issue is the willingness of people to participate in and support the IETF process and follow the recommendations of the IETF's Instant Messaging and Presence Protocol (IMPP) working group after the current phase of the work is complete.
I left the meeting feeling that people would go along with the IMPP's recommendations. Once the IMPP working group has completed its model and requirements documents, they will be focusing on the protocol discussions THAT could be lively. For most of the Telecom guys at the meeting like dynamicsoft, Ericsson and OZ.com, their proposed answer to the problem was SIP.
One of the interesting conversations was regarding the role of presence and the way it relates to initialization as defined in SIP. The discussion was that while SIP has a role and is useful when providing service, for presence management it is not necessary. We are not at the point where these ideas can really be debated, but SIP maybe the right transport for IMPP messaging, but not necessarily the answer to the question of how to transmit and view the who, what, etc. for the presence information.
As a group, the consensus was that the work in progress at the IMPP should be actively supported with both expert technical help and a commitment to push the standards and the next IMPP working group charter forward.
Having met with the industry, I believe a gathering spot for the
industry is important and to that end, I've decided to refocus our planned Directories
on the Net conference to look at Instant Messaging and the effects of being
"always on". Instant Messaging '99 will be taking place November 9-11 at the
MGM Grand Hotel in Las Vegas, NV.
The Market for Low-Density Gateways
It is becoming clear that there is definitely a market for low-density gateways
between 2 and 4 ports at a $300 price point. Embedding protocols in edge devices
is popular for immediate implementation of IP Telephony and perhaps one day
providing for IP across a large portion of the network, a la AIN. Such small
devices, apart from being nifty, are an easy way of creating hot, mid-quality,
inexpensive connections between points that still can utilize current telephones.
TIP of the Iceberg?
Trying to define a "communication portal" for the upcoming VON preconference
workshop on the topic, we decided that the term was too ambiguous, especially
in light of recent hype and the complete misnomer of a certain VoIP billing
company. Instead, it became clear that the software and websites of interest
to VON Developers were actually Talk Initiation Points (TIP's). Including directories,
interactive web pages and buddy lists, is is a category of items that encourage
a person to engage in VoIP (or other communication methods) and give them people
to talk to.
One part of this is products like Odigo and Gooey, and to a lesser extent ThirdVoice, which add external interactivity to web sites. Allowing users to interact with others on the page, they are slowly giving away to voice chat rooms where very page is a new conference. Products like Lipstream and Hearme, although they are getting there, have a long way to go to make their buddy list and browser monitoring catch up with the messaging-only clients.
A second part is the unified inboxes. These web pages or dial-in services, such as Onebox and Deltathree, allow voice mail, email, pages and fax to segue into being relayed or replied to over the phone. Products like iPulse, meanwhile, route callers into a communication method of appropriate priority, which may or may not be voice.
Another way to go is regular buddy list functionality, available in Odigo (a Pulver.com favorite), but more traditionally seen in ICQ, AIM, Yahoo!, and Excite. Already the last two incorporate voice chat natively. For Tribal Voice and other companies, the route that presented itself was a community motif. Organized into groups, these shared buddy lists of sorts allow people with similar interests to congregate (similar to the browse-along context-sensitive clients mentioned above) and then chat.
Web pages may also reach out and touch the user, as seen with
iContact, LivePerson, and its less intrusive brethren like VocalTec's Surf 'N
Call. IP-based call centers become integral here, allowing streaming voice to
be a pathway between people and companies. Aggregate directories are also a
necessary element in initiating communication. A central repository makes it
possible to locate people in order to communicate with them, and, sadly, directories
from writers of messaging clients are not up to snuff. This more passive TIP
is still better than the legacy phone book.
The Un-Controllers: Un-PBX versus the Un-Central Office
Using software to control a phone call is basic to communication systems. The
Central Office is primarily a stored control program managing IO from its ports.
But the Central Office was designed to reduce the cost of gathering customers
into one common "central" office. Enterprises could benefit independent of the
phone company. So before the Internet hit critical mass, the benefits of local
control of phone service created the PBX market. This market became the stomping
ground of the computer telephony integration companies. As the enterprise WAN
became less of a Voice VPN and more of a bridged LAN, this caused the rise of
the un-PBX. These systems were designed to look like a PBX. As the actually
wire facility has become transparent, VON became the method of service. Now
we are at a point where the benefits of a common central office (CO) aren't
as clear. And the US regulatory environment with competitive access and local
number portability is making the CO less relevant. The result is of the extraction
of the media gateway controller from the gateways themselves is the un-CO.
During the past few weeks companies have been coming to our office and briefing us on their new product releases. Some of the companies come are targeting the role of the un-PBX and others are attacking the un-CO. The Un-PBXs target companies and point to S.100, H.323 and TAPI as proof of an open standard. They are correct that in effect have an ability to provide feature manipulation to their customers through their architectures. It also represents a model of service creation well know to the IT manager, the features either come with the "gear" or you have to build using internal resources". Many issues such as traffic management and network implementation are doable within this microcosm. Likewise they benefit from the fact that the single wire concept is dead. And many of these systems gladly relinquish control to legacy systems such as PBXs.
On the Un-CO side the effort is in comparison a treaty to establish world peace. Here the target is often less focussed. While Call Agents talk with single port gateways, they also talk with next gen tandems. The model is MGCP/SIP and the web. Here the thought process is that developers will be both webmasters and HTML writers. Coding will be the responsibility of anybody looking to use features, better known as the "The self sufficient end user". (What about all of us who are members of the "Flashing 12 Club and still don't have their clocks set on their VCRs?) Think you need a manual for your cell phone now. Just wait. Wait till you see the Netfusion offers their SIP objects for call management. Of course this may not be that much of a hurry. SIP has some hurdles to overcome (even with the help of MGCP). While the SIP and Megaco camp represent their particular concerns, it is doubtful that we will see the deployment of these systems without a thorough workout in the carrier labs and some compelling reason beyond call completion. An alternate perspective is that as MCI builds new services on SIP that are so compelling, it will end up launches massive work efforts by their competitors to maintains status quo.
So the compelling reason is the features. And the SIP focus should not be on the service provider but at Barnes and Nobel and Amazon.com. The goal is to get the webmaster interested in marrying his applications to the next gen network. We need an O'Reilly animal book on his desk with a few reference samples some neat bells, like an automatic redial client on a web page for reorder and order status calls, or maybe an ability to call the helpdesk and pre-select the IVR to the right spot rather than having to listen to the menu driven crazy code reference.. As I'm writing this I can imagine doing a little CGI and having this available on my own site - Now all I need is distribution. And it is possible to imagine a Application Service Provider that provides a feature that compels people to seek them out.
However, I believe this will happen, which is why we have started
the VON Developers Conference. I believe in the future of third party development
to the point where I am willing to collaboratively fund / edit the book if we
can find SIPPers ready to provide chapters and sample code. We're also willing
to do the same for the TAPI types. Because ultimately this war is won by what
tools are available that are flexible enough for coding and useful enough for
the customer.
Neutron Mergers and Acquisitions - Cisco
A recent BusinessWeek article mentioned that Cisco's employment pages are not
intuitive for the applicant to connect themselves to the job. If I were an employment
agency, I would say the solution is simple. Don't apply at Cisco, apply at the
next or near next acquisition. You will probably make more money in the conversion
of your options then you will as a direct new hire. But don't expect that what
you produce lives on. Look at the recent history Cisco Amteva, Cerent, and now
Monterey. Who is next?? And from Cisco's perspective Don't Hire - Acquire. It's
a steal buying talent then redeploying people. If you are a target remember
if the corporate assets don't fit, people are our most important asset after
all.
The Commodization of the Minute Continues
"I wanted to Bill you, but I Realized it Would Cost Too Much" may be a common
phrase heard within the board rooms of phone companies around the world in countries
where the minute has become a commodity and there is a no visible floor for
the price of the minute.
AT&T recently announced a new calling plan in the United States,
so the price wars for Long Distance Rates within the US are moving forward.
I am happy for the consumers, but I am sure that the Billing companies are wincing
as the rational for call details in the legacy systems continue to dwindle.
You can't talk about new revenue without new services and on the vicious cycle
continues. Everyone fights on a minutes basis and the dogs are getting leaner
and hungrier. As the pricing wars continue I can imagine the pitch "commit to
500 minutes and get your toaster free". Maybe bundling will become more and
more Amway like - phase 2 cajoled friends and family. With the prices down so
low, maybe we'll get free prepaids that are filled with logos. Who knows, it
maybe thats the most profit available. Better yet go for the collectable prepaids
- get your Pokemon Prepaid Cards and if you dial this 900 number you can fight
with Pikachu at $3.00 per minute and get additional credits on the prepaid.
Making No Difference: Directories Interoperablity Forum Update
The Directories Interoperability Forum met in Atlanta September 13th to discuss
their work activity. Carl Ford, the pulver.com community developer for directories
attended the meeting and reported that the group's focus is on getting past
the current state of the art where even LDAP is not compatible between systems
and to a point where an application can be developed on one system without needing
to be rewritten when used with another system.
The directory vendors intend to support this by developing an accreditation process for them to pass which will be administered by the The Open Group. This will be a long process and requires some forward thinking by all the voting members.
One key goal is to carry the message and get feedback from the
application developers and integrate software vendors. To support this effort
we will be present looking for active discussion about the problems faced by
the development community at directory@pulver.com. Carl also will be facilitating
birds of the feather sessions at the Directories conference and at Fall VON.