I started at INT a few weeks ago and my first task as a new INT developer was to add a data connector to IVAAP, INT’s HTML5 visualization framework for upstream E&P solutions.
As a new member of the software development team, I had no prior experience with development on this platform. To gain knowledge of IVAAP and to understand more about the IVAAP software development kit, I used the IVAAP developer’s guide. I found this guide quite useful as it made the key points behind IVAAP easily understandable.
With only a few years of experience with Java, I was surprised by the lookup system. IVAAP has a microservices REST architecture and is very modular in nature, and the lookup system ties all these modules together. It’s quite powerful, but this is something I had never encountered before.
Coding with IVAAP uses a simple model where each entity implementation consists of implementing a POJO (Plain Old Java Object) class and its finder. This paradigm is consistent throughout the entire code. Essentially, for this project, I plugged only a few classes:
- A data source type class
- A data source class
- A log curve class and its finder
- A log curve data series class and its finder
- A log curve data frame class and its finder
Implementing these classes essentially consists of following templates, where the public API provides hooks and the developer adds the implementation specific to their project. The public API is documented, making it clear what each method or class is meant to perform.
Even though this project is to be deployed on Linux, my development environment was on Windows. It consisted of an Integrated Development Environment (the NetBeans IDE) and the Postman tool for testing individual services. This project accessed a SQL server database, so I used dbVisualizer to browse the data.
Since this project only included adding a data connector to IVAAP, I didn’t try to add new services, only extend the data sources it supports. Building a connector on top of the existing web services allowed me to validate my work as it progressed. For example, after plugging a new data source type, I could immediately verify that it worked as intended using Postman and following the HATEOAS links. This remained true when I plugged a data source and each finder. No need to wait until all classes are plugged to verify that the logic works. I also found that the error management built into IVAAP helped me be efficient since the error report made it easy to trace the actual issue.
The learning curve of the IVAAP software development kit is gradual. The API guides you. Unlike some of the frameworks I have worked with, there is no prior knowledge necessary to get started. You can be effective from day one with just basic Java knowledge.