Re: Re: Abbreviated keys for Datum tuplesort

On Fri, Apr 3, 2015 at 1:17 PM, Stephen Frost <sfrost@snowman.net> wrote:
>> but even I'm not willing to
>> expend the amount of ink and emotional energy you have on whether a
>> variable that holds +1, 0, or -1 ought to be declared as "int" or
>> "int32".  Does it matter?  Yeah.  Is it worth this much argument?  No.
>
> My only comment will be on this very minor aspect (and I'm quite
> agnostic as to what is decided, particularly as I haven't read the patch
> at all), but, should we consider an enum (generically) for such cases?
> If that's truly the extent of possible values, and anything else is an
> error, then at least if I was writing DDL to support this, I'd use an
> enum, maybe a domain, or a CHECK constraint (though I'd likely feel
> better about the enum or domain approach).

It isn't the case that an enum can support it. Some comparators will
return -5 rather than -1 (as with C-style comparators in general,
sometimes comparisons can be implemented using subtraction and things
like that).

-- 
Peter Geoghegan