Andrew Pimlott wrote:
> On Wed, May 08, 2002 at 06:47:46PM +0200, Zeugswetter Andreas SB SD wrote:
> > When we are talking about the places where you need double escaping
> > (once for parser, once for input function) to make it work, I would also
> > say that that is very cumbersome (not broken, since it is thus documented) :-)
> > I would also default to strict ANSI, but not depricate the escaping when set.
> > All imho of course.
>
> As the original reporter of this issue, I am gratified to hear it
> acknowledged by the developers. Thanks! (I also apologize if I
> exaggerated the pain caused, as apparently not many other people
> have been bitten by this specific problem. Well, it was painful for
> me. ;-) )
>
> I must say, though, that I remain bothered by the "not broken"
> attitude. There is an obvious standard for PostgreSQL to follow,
> yet it is non-compliant in utterly trivial ways, which provide
> marginal or no benefits. Granted, changing long-standing defaults
> may not be acceptible; but there is a big difference between, "it is
> broken but we just can't change it for compatibility reasons", and,
> "it is not broken".
>
> It is my experience that most other free software projects take
> standards compliance more seriously than PostgreSQL, and my strong
> opinion that both the project and its users (not to mention the
> whole SQL database industry, eventually) would benefit from better
> support for the SQL standard.
>
> Ok, I've said my peace.
Yes, these are good points. Our big problem is that we use backslash
for two things, one for escaping single quotes and for escaping standard
C characters, like \n. While we can use the standard-supported '' to
insert single quotes, what should we do with \n? The problem is
switching to standard ANSI solution reduces our functionality.
-- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610)
853-3000+ If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill,
Pennsylvania19026