Re: PostgreSQL 10 parenthesized single-column updates can produce errors

"David G. Johnston" <david(dot)g(dot)johnston(at)gmail(dot)com> writes:
> ​Definitely moderates my opinion in my concurrent emai​l...though
> postponement is not strictly bad given the seeming frequency of the
> existing problematic syntax in the wild already.

Yeah, I'd hoped to get some capability extensions done here before v10
shipped, in line with the theory I've expressed in the past that it's
better if you can point to actual new features justifying a compatibility
break. However, that didn't happen in time.

I'm disinclined to revert the change though; if there are people making
use of this oddity now, then the longer we leave it in place the more
people are going to be hurt when we do break it.

If I had a time machine, I'd go back and fix the original multi-column
SET patch so that it required the word ROW in all cases --- then at
least it'd have accepted a conformant subset of the standard syntax.
Oh well.

regards, tom lane