Some nice information on the “In Memory SRF” concept incorporated in IP2015, this will get its final shape in IP2016….!!
SRF as such is not going away. We are only changing how we store the output of the compilation process from a file to tables in the repository. When the server is running in Composer mode, it loads the repository information from the repository tables, rather than a file, thus allowing us to publish new content while the server is still running and the next session that starts up will load the new content. Since repository information is loaded & cached once at login, active sessions are not impacted by new content that is published in the background.
Publishing new content uses the same compilation process as before, with the only difference that we now store in repository tables. This reduces the risk of migration from SRF to In Memory SRF as the core processes are still the same.
Siebel CRM Composer enables a set of new features that are not available in SRF mode. For e.g. Server side Responsive Web Design is achieved by removing the need for file based Web Templates. This requires additional repository schema to hold additional data, which is read and processed by Siebel CRM Composer only. SRF is unaware of this data even though it resides in the same underlying tables. The Schema model for Siebel CRM Composer features are designed in a way that they can coexists with Tools. However, only one mode can be active at a time.