Over the past several months--several years, actually--I’ve found myself asking the following questions innumerable times, sometimes to myself and my associates and oftentimes to high-tech industry players of all stripes:

• “I understand what you say you are doing, but how do you connect your VoIP subscribers to VoIP subscribers on another provider’s network without using the PSTN?”
• “Why do you have to convert and re-convert traffic from IP to TDM to IP again several times? Why can’t you just keep the traffic as IP from caller to receiver?”
• “Why does the call have to go through so many codecs? Will the sound quality really get a MOS score of 5?
• Why won’t my mobile phone work as a modem for my wireless-capable PC or PDA? Why do I need a modem anyway?”
• “How are you going to be sure that your business partners, maybe on other continents using different network providers, will be able to join your multimedia collaboration session?”
• “Why can’t I use my GSM phone on any mobile network providers network?”
• “Does the Brand X decision mean that you might block someone using Vonage or Skype for telephone calls when they use your broadband access service?”
• “How are you going to handle address translation and directory look-up when the called party is part of another enterprise network?”
• “Why won’t my email to someone@ISP.au go through on my ISP, unless I use someone@ISP.au.com as the email address?
• “How will you verify that you and your casual content partners are sharing revenues correctly when you sell their ringtone or movie or other content to a subscriber roaming on your network?”
• “Why does the hotel block my enterprise-hosted e-mails but allow my Web-based e-mails when I connect via the broadband connection in my room? Why does it allow my enterprise-hosted e-mails when I use the Wi-Fi in the lobby?”
• Why does my TiVo always have to be connected to a phone line?

In most cases, the answer I received was less than truly gratifying. The immediate problem or issue may be resolved, but the solution is generally short lived and only resolved for the walled garden inside of which it was asked (for example, “We can let you connect to your office email from any of our hotels if you do this--but the solution won’t work next week in another hotel chain’s room”).

What do all of these gnarly questions have in common? Why must they be asked? At all? And is there a likely solution (or set of solutions) today, or in the offing?

All of the questions speak to interoperability of one form or another: protocol, speed or address translation; network (access, edge or core) incompatibility; privacy and security protection; peering or interconnection arrangement; or, finally, as-yet-unresolved business model concerns such as making a return on an infrastructure investment of billions of dollars.

And, there is likely at least one--or, more likely, more than one--answer to each question or way to solve the issue. And there are probably almost as many schools of thought about how the question should really be answered, as there are questions; or maybe even more. Let’s look at answers to these questions from two extreme positions--the smart device on the dumb network and the smart network supporting the dumb device.

One school of thought is the smart device school. Make the device capable of connecting to any network when it is turned on (and plugged in, if required). And once connected to a network, let the device determine what baseline capabilities the network provides and how it should accomplish the task the user wants over the dumb network to connect to another smart device somewhere.

Another school of thought is the smart network school. Find out what devices are connected and give each device the capabilities it is allowed to have, based upon the rules the network has in its database, for that device.

Neither of these polar schools answers the real issue of connecting users to users to enable a communication, not just devices, but adding user identity information into the device coupled with user subscription information in the network (someplace) is required to solve even this basic fundamental issue.

In our world or ever increasing complexity--more and more devices and more and more capabilities--neither polar school will graduate a student with a perfect solution and answer to all of the interoperability or peering questions alone. But there is one thing I am certain of: If both schools don’t learn to work together, we will wind up with more walled gardens where the users of these increasingly smart devices will be limited on how, where and when they can connect to other users and, no matter what the business model, the return on investment of the device, network, service and applications providers will all be sub-optimized. So, “Why can’t we all get along?"

David H. Yedwab is the Executive Vice President of The Eastern Management Group and can be reached at dyedwab@easternmanagement.com.