In general, there is no need to generate an RMA from the command line. Generally, an RMA only needs to be generated for a project once. After that, refreshing from the repository will fully update the RMA. Then the code generated for the RMA can be copied from machine to machine as you migrate to production. The main reasons I have seen to regenerate an RMA are if the options change (such as table of contents or signon), the Include/Excludes change for what is included in the RMA, or if you have "released" the project in Blaze. With some study, all of these items can also be modified in the generated code.
Is there a specific reason you need to generate the RMA from the command line?
Our CI server currently spits out ADBs as changes to the Blaze repository are committed. I would like to generate the RMA from the command line so it too can be integrated into the CI build process. This way, humans don't need to decide if a new RMA is the result of repository changes, the CI server decides.
The other way of looking at it from a SCM point-of-view is: building any artifact from a desktop machine using human hands is inherently lame.
The RMA adapts to changes in the repository and reflect them. In order to refresh your session (in case of concurrent modifications), you can simply use the refresh (F5 or CTRL-R) command in your browser. If you are disconnected, those changes will obviously be taken into account when you log in.
Your RMA will reflect:
- changes in templates
- changes in instances
After your very first generation of an RMA, you do not need to generate it ever again (unless you want to change some configuration). No need for a lame code generation
VP, Product Management
Retrieving data ...