Field and lookup changes in the RESO Data Dictionary can create pain points for MLS vendors and customers, especially when your MLS committee demands that “Distance to Volcano” should immediately be added as a field in the MLS.
During the 2022 RESO Spring Conference, Paul Hethmon, CTO of AMP Systems, explained how a complete embrace of the Data Dictionary has allowed them to build a system that can make changes in minutes, not months. | WATCH VIDEO (9:15 minutes)
The solution is born in the new Field and Lookup resources for core metadata definitions, enabling the creation of a highly flexible system for MLSs.
The Data Dictionary is made up of fields and lookups, which cover a lot of ground but not all of it. However, since the dictionary is extensible, an MLS system can create local fields for what the Data Dictionary does not have (e.g., Canadians encounter Property Class instead of Property Type and Residential Freehold instead of Residential).
Said Hethmon, “If you think about the solution that you want and not the tools you’re using today, the Data Dictionary actually gives you a great foundation.”
Going a step further beyond RESO’s T-RETS Extinction campaign, Hethmon suggested we should also think about T-RDBMS, the extinction of relational database management systems.
“When your solution is a relational database model, you tend to take your problem and shoehorn it in there,” said Hethmon.
Using the “Distance to Volcano” example, Hethmon said that adding a new column of information to a database with millions of rows could conceivably take down the system for a weekend in order to tweak the data access model, business layer, user interface and rules engine, among other needs.
With that in mind, AMP Systems asked themselves two fresh questions to account for changes:
- What if we treat fields as configuration?
- What if the Data Dictionary was our blueprint?
In this new model, fields are not defined in code or as database columns; they are metadata that drives the system and makes it better.
Having a data store as a relational database hybrid allows new fields to be placed in a blob (binary large object) storage that could be, for example, a JSON blob, MongoDB or Elasticsearch cluster, making formerly time-consuming requests much easier to perform.
