By Dennis D. McDonald

I've been thinking some more about the issues raised in my earlier posting The Inevitablility of "Too Many Gateways".

One of the reasons the situation exists (as Ismael Ghalimi described where he has to maintain many separate accounts to manage different internet gateway services for feeds and data exchange) is that it's becoming increasingly possible to create such services and to make then available on the web.

This may be related to the phenomenon described by Dion Hinchcliffe in The SOA with reach: Web-Oriented Architecture. To oversimplify Hinchcliffe, the tools are out there to create this proliferation of products and services. As long as barriers for creating systems are low and a perceived need among users exists, applications will be created by someone who has an innovative idea and whose business model seems to be supported.

The fact that so many of these systems are still in "beta" is probably a side issue, as I discussed here. Beta or not, these systems will need to be supported. This brings with it requirements for a certain level of support, bug tracking, upgrades, modifications, enhancements, fixes -- and version control.

If someone's business model depends upon the use by paying customers or users of the appropriate version of the application, and the storage of certain user or transaction details beyond passively gathered cookies, a requirement for user registration seems inevitable. Hence we have what can be viewed as a "proliferation of platforms" and resulting complaints by power users like Ghalimi about having to maintain and track passwords and registration details.

What is the best way to impose order on what some might view as an increasingly chaotic situation?

One approach is simple: hands off; let the market and economic Darwinism decide. Sooner or later people have to start asking themselves, "Do I really need to sign up for another RSS based [fill in the blank] service that I'll have to take time to read and digest?" People do vote with their feet, after all.

A second approach is a voluntary grassroots move towards a common platform for registering and managing preferences and privileges. Let's assume that is technically possible. I have questions about the business model for such an approach and whether it is politically feasible. For example, there may be a limit to how far the open source model can be taken in the commercial marketplace without an explicitly identified revenue source. Also, a "common platform" for managing what might be viewed by some as an essentially administrative or overhead function might not be viewed as a hot prospect for generating ad-based revenue.

A third approach is that the "big guys" will negotiate and/or decree certain platform guidelines for handling multiple registrations and basic registration information. They will then incorporate these guidelines in their offerings to risk-averse corporate clients and to the public. Business will then go on as usual, with the innovative "little guy" getting lost in the dust (unless he or she is willing to adopt the new "standard"),

A corollary to the third approach is government intervention and the imposition of regulatory standards in certain areas where interchangeability of certain types of registration data is mandated in order to maintain, for example, a level of access equality with regard to certain types of products and services (e.g., those related to medical service, health and safety, or personal security).

What do you think?

Let me know, either with an email to ddmcd@yahoo.com or by using the comment field below!