On Thu, Feb 13, 2003 at 09:39:15AM +0000, Dominic Mitchell wrote: > Rudy Lippan wrote: > >On Wed, 12 Feb 2003, David Wheeler wrote: > >>I think that Tim's idea was for a new DBI attribute, CharSet: > >> > >> http:[EMAIL PROTECTED]/msg15486.html > > > >I am game with that. But I'd still like to do some research on the issue, > >and I would also like make sure that DBD::Pg goes they way DBI goes. So I > >don't want to rush and end up with a bad interface or one that differs > >from the other DBDs, if possible.
Drivers are free to add whatever they need to solve their particular problems so long as new attributes and methods begin with the driver prefix (pg_ in this case). It's often the case that one or more drivers implement private solutions to an issue before the DBI adopts a common API. That's a good thing as it helps ensure that details have been thought through in fully functional implementations. (As per the IETF motto "rough consensus and *working* code".) > There are lots of semantics to decide with regards to character encoding > in databases. And it would be fantastic to have a consistent interface > as part of DBI. One of the first jobs is to see what kind of support is > available from different databases and try to come up with a common > interface... Is anyone willing to champion this work? Tim.
