Hi, Last night I made a big improvement on the generator-polygene, making a lot easier to work with the codebase (modular, creation-by-convention,...). Still room for improvement in that area, but it is now miles better.
I am toying with the idea of generating ALL possible projects in the integration tests, to ensure that no conflict exists, or that something is missed in some combination. But my Groovy-Fu is just not "there", and I don't want to look "silly" by having a Java program with Groovy syntax. In essence, generate 2688 JSON documents with all possible combinations, and pass them one by one to a command line. Why so many? 1 of 14 entitiy stores x 1 of 4 indexing x 1 of 2 serialization x 1 of 3 caching x 1 of 2 metrics x each of 2 features (4 combos) And with each other feature I add, the number doubles, so perhaps not doable in the long run, as I intend to include the following features eventually; 'rest api' ,'security' // ,'version migration' // ,'logging' // ,'jmx' // ,'circuit breakers' // ,'file transactions' // ,'spring integration' // ,'servlet deployment' // ,'osgi support' // ,'alarms' // ,'scheduling' // ,'groovy mixins' // ,'javascript mixins' Other things that I am working on are; * Support for adding Domain Layer modules with entities, values, objects.... * Support for different application types; Rest, command line, WAR, Docker(?) * Post creation-time, creation of new modules with entities, values, objects.... * and hopefully, post creation-time addition of features. The last two are of course very tricky, since humans may have been tinkering with files, so perhaps that is a pipe-dream. Anyway, if anyone could help out with the tests, then this could be wrapped up pretty quickly. Meanwhile, I will continue on the other features. Cheers -- Niclas Hedhman, Software Developer http://polygene.apache.org <http://zest.apache.org> - New Energy for Java
