Derek Baker Sunrise Technologies, Inc.

We really need more control over the true schemas used by SSIS in data management.  We need the ability to view as well as control the schemas further.  We have had a variety of situations throughout the process of the project implementation and in production where data management recurring integrations will fail and we find out that SSIS is complaining about schema validations.  But we cannot verify this to be accurate or not since we have no access to the schema.

As an example, we had been integrating a composite entity file with a single child entity for some time.  Sometimes, the file would have the child entity and sometimes it would not.  This was fine but when we applied Platform Update 5, this ceased to work and would only process the file if it had at least once instance of the entity.  This certainly has to be something that can be controlled in a schema file and was something we at least could view in AIF in 2012.  Please provide more visibility and control into the schema files used by SSIS

Currently, OData supports atomicity strictly with a single entity or by the use of the $batch service.  The $batch service is great and has it's purpose but is very burdensome to implement.  A very common situation is to post a document (i.e. a sales order) with a header and line structure but require that it be posted in an atomic manner.  This can be done with the $batch service but only with a lot of additional work.  If we could use deep inserts to post the header and lines nested within all in the same operation without the need for additional overhead of the $batch service, this would be a huge gain and would get us much closer to what the AIF had available to perform.