July 31, 2010

Daunting Details

Accessing the electron pattern references became a standard practice. It was not something that we had invented in the electron network, but it was something that we had borrowed and improved upon. The central network, while simple on the surface, was more complex than the electron network that we were constructing and building out. Where possible, we borrowed structures from the central network, but in cases where the signals became distorted, we were left to our own devices.

Differential frequency encoding worked for the patterns that we encountered on a regular basis. The oddball patterns required disassembly and re-construction to understand the fundamental differences and structures that existed in these nuanced patterns. Careful observations of the local response to some of these patterns revealed that it was possible to come up with a reasonable token encoding for the pattern, but for low-use patterns, a single pinger would suffice to cover the entire set of related patterns and provide a destination location for creating the associated triple-tap token.

With the final deployments, triple-tap encoding system came online as a way to quickly reference new patterns and zones in the central network. As these patterns repeated and demonstrated higher frequency of use, we assigned more appropriate tokens to them. Overlapping pingers were appearing now and again, which usually combined into a single token, as the concepts that were embodies in an overlapping situation were closely related. Here, we were finally discovering some of the meaning assigned to the tokens. I was just happy that the parser was working and that the lookups occurred.

No comments: