SUBSCRIBE TO TMCnet
TMCnet - World's Largest Communications and Technology Community

CHANNEL BY TOPICS


QUICK LINKS




SOA Web Services

SOA/Web Services

 

SOA/WEB SERVICES FEATURE ARTICLES


February 10, 2006

The Death of the Computer Telephony Interface

By TMCnet Special Guest


By BRIAN SILVER
TMCnet Special Guest
 
PBX CTI has passed on, and I’m not sure anyone noticed. Hanging on for dear life, too, are the TAPI interfaces. Even desktop operating systems come complete with a SIP and RTP implementation that’s surprisingly standards compliant. We’re all familiar with the advertisements that hawk the phone as the central element of the desktop – it’s presented as your office, the internet and world-wide-web and phone service all wrapped into one. PBX CTI and TAPI make this implementation possible. But, last I checked, that poor phone is anchored to the desk and isn’t nearly as mobile as a laptop PC or aptly named mobile-phone. The “mainframe & terminal” model of the PBX and its resulting programming interfaces just don’t scale in a nomadic and always-connected world.
 
All of this is kind of sad in a way. The idea of tying together computer programs and telecom information systems is powerful. But then again most mainframe programs were powerful too, but the flexibility and cost reduction that came with the “PC Revolution” enabled the even more powerful Internet technology we use today. Just like the other mainframe systems of the past, the PBX will never die. It will change its spots and become a server for those specific applications that it will remain good at.
 
Strangely as we move from the desktop-centric model of telecommunications to a model that mirrors the distributed, networked application world, the ability to interwork data applications and telecom applications becomes even more powerful. For example, BEA recently announced a call-center initiative that relies on WebLogic - BEA’s application integration platform for service oriented architectures. Pretty cool, huh? Call-center software no longer needs to be in the “center” of a contact-management strategy. Moving to an SOA model of call centers allows the enterprise to remain extremely flexible on outsourcing-and-then-insourcing-and-outsourcing-again.  While there’s power in this model – great power in fact – there’s a shortcoming. How does this all interact with the telecom infrastructure?
 
Hook the whole application integration environment to the PBX, right? The PBX was built to enable telecommunications to the desktop, but the SOA enables applications to completely forgo any knowledge of old enterprise boundaries. What does it cost to enable cellular-reachability via the PBX so that the new applications can have limitless geographical reach for all interactive communication? What about leveraging the Internet as another globally-addressable telecom facility? PBXs are notoriously bad at problems that involve nomadic use and, combined with the PSTN, create expensive solutions for any communication that involves customer-to-customer, customer-to-partner communication, or requires any form of external dynamic call control outside of the traditional enterprise boundaries. The PBX architecture and cost structure virtually eliminates the gains made by using the SOA to create interactive business processes. Sheesh, you could have used the old CTI interfaces if this was the desired outcome.
 
In other words, a quick look at what people want to do with application-integrated telecom tells the story of the PBX CTI demise. The web-service and SOA worlds are making strides that allow the enterprise to rely on more external services to implement their business processes that will be used by their users in locations that only require Internet connectivity. Talking, listening and seeing are big parts of most business processes and certainly need to be incorporated seamlessly into the enterprise application architecture. Telecom service peering, partner federation (i.e., extranets) and technology abstraction (that is, Cellular vs. PSTN vs. VoIP vs. PBX) need to be integral and transparent in the new telecom integration paradigm just as other technologies are generalized in the application integration architecture.
 
So is my report on the death of the PBX CTI grossly exaggerated? Depends on how you look at it. Companies like BEA are doing some heavy lifting here, but without the further integration of the telecom capabilities deeper into the SOA architecture, the silo effect of the mainframe-PBX will continue to hamper their efforts and web-service interfaces to the PBX will be nothing more than “tomorrow’s CTI.” Forward looking vendors will start to incorporate the telecom capabilities – call routing, transport functionality, and services like conferencing, voicemail and music-on-hold – into the application server itself to help mitigate this problem.
 
----
 
Brian Silver is chief technology officer with BlueNote Networks, which is pioneering the collision of SIP-based interactive communications with service oriented architectures (SOAs) by delivering the first enterprise-class interactive communication platform: SessionSuite. He can be reached at [email protected]

SOA/WEB SERVICES





Technology Marketing Corporation

2 Trap Falls Road Suite 106, Shelton, CT 06484 USA
Ph: +1-203-852-6800, 800-243-6002

General comments: [email protected].
Comments about this site: [email protected].

STAY CURRENT YOUR WAY

© 2024 Technology Marketing Corporation. All rights reserved | Privacy Policy