Content provided in partnership with

Technology Industry

The latest, greatest computing architecture - Internet/Web/Online Service Information

Communications News, August, 2000 by Bart Taylor, Dan Taylor

There are so many smoke-and-mirror images in the current market for application services, that it is sometimes difficult to convince people we meet for the first time that we chose to "cover" ASPs and the infrastructure to support them. This is especially true because ASP is a development-stage market with scant numbers of customers for "pure play" ASPs, delivering enterprise applications as managed services.

Why should we all care about ASP? Computing architectures are evolving toward greater and more flexible support for networked application services. As this happens, the shape of new service infrastructures will begin to more closely match the computing architecture of the future enterprise. This is a significant new development that ties the fates of services and enterprises closely together.

Most Popular Articles in Technology: An overview of continuous data protection; Why all those current ratings?; Many countries now have a mobile penetration rate above 100%, report says; The Tata Group's big telecom gamble: VSNL's recent acquisition of Tyco ...; MEASURING BANK BRANCH EFFICIENCY USING DATA ENVELOPMENT ANALYSIS: MANAGERIAL ...; More »

Last month, we described the debate between proponents of client/server and Web application server architectures. We tilted at the idea that Web application servers undo the shackles of bad old "legacy" computing. This month, we would like to opine that n-tier product developments in "thin client" technology overcome at least some of the working distinctions between client/server and Web application servers.

Giotto Perspectives covers the service-provider environment. In some respects, new Internet application technologies do provide tremendous potential for flexible integration and the creation of new services. In our past months' columns, we've backed this very idea, and we continue to see rapid deployments of new architectures by service providers in particular.

But let's dispel some myths here. Webification of apps is indeed important, but it is not a magic cure. First, there is the assertion we've sometimes heard that Web architectures are inherently more scalable. We talk with service providers of various types for a living, and most of them warn us away from this "just-add-architecture" theory.

A computing architecture alone probably does not future-proof a real-world provider against requirements to scale. In particular, customization and integration requirements are not generic, and therefore even when controlled within a service environment, these can introduce the same sort of nasty surprises that client/server implementations have seen. This is a reality both enterprises and service providers share, but to some extent service providers may suffer more harsh consequences from unforeseen costs.

Second, there is a widely held belief that new HTML variants, as well as .asp and Java front ends are overcoming the key limitations of the user interface that browsers and Internet connections currently provide. With a rich new user interface, Internet applications would gain a real purchase in the war against client/server applications. Possibly the new browsers will do this, but what service provider is currently banking on this?

In some contexts, such as wireless application gateways, diminished-capacity interfaces offer real value. Otherwise, the interface needs to improve, and if the history of commercially successful data services is any indicator, this won't happen overnight.

On the client/server side, the argument is more grounded in market presence, including well-developed channels, established brands and erstwhile war chests drawn from existing revenue streams. This market position is nothing to sneeze at. In the world of start-up ASPs, there are many companies intent on blazing a trail using good old client/server applications. Many of the client/server software vendors have moved ahead in recent years with their own Web application server migrations, but again, this is not happening without delay, and the hazards of integration still exist.

We are critical of this "us vs. them" mentality about architectures, in part because our research conversations with service providers reveal few compelling reasons for it. A fractious debate strikes us as being less and less relevant.

Over a year ago, the only discernible set of ASPs appeared to have emerged from either outsourcing or integration backgrounds--and a surprising number of the integrators-turned-ASP started out with Citrix expertise. Without a doubt, the established presence of the Citrix Metaframe product has paved the way for ASPs, including Citrix channel supports for VARs and others who want to become service providers, ASP-style. Giotto Perspectives has lost track of the sheer number of ASPs who come from the Citrix camp, but it's impressive.

What is most important today, though, is the exciting new developments on the product side. Citrix has added new product capabilities that extend Web delivery capability for client/ server applications and greatly enrich user-interface development for applications to heterogeneous device types, to more closely meet the expectations of individual users. Tarantella, a competing product, was offered this year by SCO, with some fanfare surrounding the incorporation of a Tarantella ASP edition into Nortel's Pre side ASP platform.

Technology Industry

The latest, greatest computing architecture - Internet/Web/Online Service Information

Communications News, August, 2000 by Bart Taylor, Dan Taylor

Find Research Guides for:

Find Featured Titles for: Technology

Most Popular Articles in Technology

Most Popular Publications in Technology