...
| Info |
|---|
Before the code base was moved to git, the platform repository in svn had the complete code base of all the products together. Thereby, releasing individual products using this svn-based model was done in chunks. A chunk release consists of one or more product releases. Therefore, with every new chunk release, the individual components that goes into the products in that chunk release were branched internally using directory structures, and then released using separate pom files that groups group all the modules (service-stubs, components, features). |
When the source code was moved from svn to github, the code base that was in svn had to be refactored into separate individual repositories. With git, everything is based on branches. So the above branching strategy (in svn) using directories will not work. Therefore, the chunk based release of products has to be changed. For this reason, the platform repository was divided into multiple, individual git repositories by grouping common components. For example, carbon-deployment repository has grouped all components (including service stubs, components and features) related to the deployment of artifacts (webapps, axis2services). These individual repositories can then be released separately and the products depending on these repositories will be released afterwards.
...