[ https://issues.redhat.com/browse/TEIIDSB-197?page=com.atlassian.jira.plug... ] Ramesh Reddy commented on TEIIDSB-197: -------------------------------------- I am not sure why you say it is an issue? it is externalized and fully automated, and in absence of missing configuration, it works exactly as you describe minus the driver versions (which I agree to needs improvement here, to remove driver versions from Operator). Another reason the driver versions designed this way was to override them if user wanted to use a different version. I am reluctant to create 30 more projects with a single file to remove one config file. The dependency part is attractive, but I have yet find solution for replacement with different versions. This is also going to be used probably with other toolings like vscode plugin, where I need types and properties. Pretty soon, I also would like to add property names and then validate them when the CR is provided Another issue I saw with Syndesis is, in Syndesis I was using URL sniffing to come up with what kind of source it is, I could easily add to this and provide that ability in the future, so it gives that flexibility rather than hide only behind the name-based patterns. Agree that, what I am saying even I have blank json file, it works as you say. This file iterates the same or lets you overwrite if you need to, for example, I used this for "soap" and "ws" where use prefix of soap for example. It gives flexibility and a single contract between the tools. How should pass the configuration? I completely miss the point you are making with enviroment-post-processor, are saying no prefix needed at all? How am I passing the properties I configured for a data source to a SB process? even if it is, why add more complexity? -- This message was sent by Atlassian Jira (v7.13.8#713008)