On 28-6-2014 11:14, James Starkey wrote: > You shouldn't have to care. The status vector is a well defined, self > describing data structure. A transmission protocol should pass it, not > interpret it. If you have to get involved in the semantics, you're > either doing something wrong (or somebody screwed up the architecture).
For clarification: I am working on the protocol implementation in Jaybird (the JDBC driver). I have to convert it to a JDBC SQLException, and to be able to create meaningful exceptions I need to interpret it to get a specific error. Lets be honest, the isc_dsql_error error code is next to useless as an error code, it provides almost no information. > That said, the philosophy is that the frst code is primary -- something > an application programmer can make part of his application logic. The > secondary codes are the details which are likely to be implementation > and version specific. Secondary codes are most useful for debugging and > unexpected error diagnosis. I disagree: some of the primary codes are low on information, and some of the secondary codes actually provide the information you want to have to be able to transform the status vector to a meaningful error. Mark -- Mark Rotteveel ------------------------------------------------------------------------------ Open source business process management suite built on Java and Eclipse Turn processes into business applications with Bonita BPM Community Edition Quickly connect people, data, and systems into organized workflows Winner of BOSSIE, CODIE, OW2 and Gartner awards http://p.sf.net/sfu/Bonitasoft Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
