When Dave assigns a job to a technician, the SData tool issues a MERGE (an HTTP POST with upsert semantics). The server updates the job, and the tool automatically invalidates the cache for that job ID.
It doesn't try to solve every problem. It solves the problem: sdata tool
When the technician’s van app syncs two minutes later, the tool requests /sdata/crm/jobs?$syncDigest=2023-10-27T15:30:00Z . The server replies: "Job 456 changed." The tool fetches just that one record. The technician sees the change instantly, using 1kb of data instead of 5mb. Critics argue that SData is "too verbose" (Atom/XML heavy) and that its query syntax is proprietary. They are right—if you are building a public API for a mobile app with five tables. When Dave assigns a job to a technician,
When Dave assigns a job to a technician, the SData tool issues a MERGE (an HTTP POST with upsert semantics). The server updates the job, and the tool automatically invalidates the cache for that job ID.
It doesn't try to solve every problem. It solves the problem:
When the technician’s van app syncs two minutes later, the tool requests /sdata/crm/jobs?$syncDigest=2023-10-27T15:30:00Z . The server replies: "Job 456 changed." The tool fetches just that one record. The technician sees the change instantly, using 1kb of data instead of 5mb. Critics argue that SData is "too verbose" (Atom/XML heavy) and that its query syntax is proprietary. They are right—if you are building a public API for a mobile app with five tables.