IBM Connect 2017: My Sessions and Documentation Constructive Criticism Call-to-Arms
Last week I delivered two sessions at IBM Connect, both with strong speakers from the community.
The first session on Tuesday afternoon was GraphQL 101 with Christian Guedemann. Within an hour of reading the documentation about GraphQL, it was very apparent to me that GraphQL was a major technology for application development. And its inclusion in Watson Workspace and Watson Work Services made this conference the perfect time to introduce it to the community. It was also a good time because Christian and I have been working on a Java SDK for Watson Work Services including full documentation on OpenNTF Wiki, which includes builders for the GraphQL queries. The conference further revealed GraphQL will be used for Connections Pink, so it’s not just a flash-in-the-pan, it will continue to be relevant beyond this conference. Philippe Riand has also integrated a schema-less GraphQL implementation for Darwino. But don’t worry if you missed the session, I’m sure I’ll be presenting on GraphQL again this year.
ODA: Super-Charging Domino Development
The second session was with Stephan Wissel on OpenNTF Domino API. Stephan has been working with ODA for some months now and is a well-seasoned speaker, so it was a pleasure to deliver with him. ODA has been around for some time, but in the build-up to Connect I completed two adjacent projects.
The first was further work on an ODA Demo Database, built as an OSGi plugin web app. The framework used – Vaadin – may not be a front-end framework I use as heavily in the future. But the approach – a web application running in an OSGi plugin on Domino – may have some relevance. This may sound scary, but it’s no different from installing any other OSGi plugin (aka extension library).
The second was an ODA REST Servlet starter plugin with full documentation on OpenNTF Wiki. REST access to Domino data was a popular topic of IBM Connect and I’ve had quite a bit of experience over the last year of creating REST servlets. Using the ODA starter gives greater flexibility for data access, as well as allowing developers to leverage all of ODA. It also gives examples of Postman tests and a Swagger file. The code also includes examples for using some core IBM libraries that make handling of web responses much easier. Of course, if you’re using the GraphNSF functionality in ODA, you get a full REST API out-of-the-box.
Call To Arms
I’ve spent quite some time on the documentation of Java SDK for Watson Work Services and the ODA demo REST Servlet. I have aimed at developing it as a best practice for documenting an OpenNTF project. The Watson Work Services one has taken into account the strengths of the LotusScript documentation, in having examples of implementations. So I want any feedback on areas it’s lacking or could be improved. Let’s create a blueprint for the kind of documentation for the next generation of OpenNTF projects.