Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> One other point worth making is that we don't always consider SQL
> compliance to be a hard requirement that trumps every other
> consideration.
Point noted.
> An example is case-folding of identifiers; it's been pretty well
> agreed that between readability and backwards-compatibility
> considerations, we simply aren't going to switch over to doing it
> exactly like the spec.
I've left that one alone both because I saw the discussions of it and
because our framework automatically fixes identifier capitalization
based on our metadata and then wraps the all identifiers in quotes.
As long as PostgreSQL honors the rules about quoted identifiers, we're
golden. :-) (Although, one of these days I may try to do something
about how psql, etc. treat identifiers which are all lower case. The
refusal to consider the quoted form a match during tab-completion, for
example, is a regular annoyance.)
> So any proposed tweaks in this area would be considered as tradeoffs
> between better spec compliance and other goals.
Fair enough. I consider myself warned. ;-)
-Kevin