-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
> it could actually fulfil the goal of DBI - being able to write > Database-driven applications that are truly independant of the database > product in use. DBI v1 cuts the porting time down to next to nothing; > but we can get it even closer! I don't think the goal of DBI is to acheive total database independence. I'm not entirely convinced that applications need to be totally independent of the underlying database: at the least, the main SQL abstraction layer needs to happen at some other layer than the DBI. I can't ever imagine rewriting my SQL statements (some of which are very complex and yes, customized for a particular RDBMS backend) to fit into some abstract description language. Nor would I trust some program (DBI or otherwise) to have the AI necessary to port my SQL statements from one database to another. It might mostly work, but it would certainly not be optimized. If I build an app that needs to run against different backends, I hard-code the SQL differences into the app. Perhaps if every DB was SQL-1999 compliant, that would not be necessary, but SQL compliance is a goal of very few RDBMS's these days. :) In short, even if it is possible (which seems unlikely), DBI is not the place for it. - -- Greg Sabino Mullane [EMAIL PROTECTED] PGP Key: 0x14964AC8 200507141317 https://www.biglumber.com/x/web?pk=2529DF6AB8F79407E94445B4BC9B906714964AC8 -----BEGIN PGP SIGNATURE----- iEYEARECAAYFAkLWn6UACgkQvJuQZxSWSsilpwCdFLFA9IVWpPpraEzcFy4QXEmQ D00AoKIxZ6xujfoLEonJaP9YtGOBxZZa =zmTS -----END PGP SIGNATURE-----
