|
In addition to a superior technical architecture, Adaptris’ philosophy to building “adapters” is very different from that of the traditional EAI adapter vendors. This philosophically different and pragmatic approach came from bitter first-hand experience of working with EAI adapters.
Having spent years working with traditional EAI adapters and being forced to write expensive legacy code to simply interface with what should have been a code-free, configurable adapter, Adaptris felt that there had to be another way. Adaptris realised that it is not possible to build a single “adapter” that can be configured to handle all possible use cases. At a certain point, a configuration threshold is reached where configuration alone no longer works and instead some level of coding is required. Most traditional EAI adapters ignore this, resulting in the ironic situation where the legacy or target system often has to be “adapted” just to make the adapter work. Add into the equation the already high purchase price of traditional EAI adapters and you start to see the fundamental flaw in their approach.
This “outside code”, or software that is developed outside of the adapter, often in old legacy technology is very difficult and costly to build. Adaptris realised that if you architected the “adapter” so that it was designed to be easily changed and extended, then more integration tasks could be solved through building the less expensive “inside code”. This “inside code”, as the name suggests, is code written inside the adapter and is much simpler and less expensive to produce and maintain.
Adaptris doesn’t advocate writing code every time an adapter is deployed, on the contrary, the belief is that all adapters should be 100% configurable and where possible integration should be code free. The difference is that pragmatism has taught that when configuration alone doesn’t work, it is the “adapter” that should change by developing inside code and not expensive outside code, and even then the coding required is kept to an absolute minimum by reusing pre-built libraries.
|