Re: Problems with question marks in operators (JDBC, ECPG, ...)

On Fri, May 15, 2015 at 1:23 PM, Dave Cramer <pg(at)fastcrypt(dot)com> wrote:

>
>
>
> On 15 May 2015 at 16:21, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
>
>> On Fri, May 15, 2015 at 4:13 PM, Dave Cramer <pg(at)fastcrypt(dot)com> wrote:
>> > Not sure what the point of this is: as you indicated the ship has
>> sailed so
>> > to speak
>>
>> Well, if we were to agree this was a problem, we could introduce new,
>> less-problematic operator names and then eventually deprecate the old
>> ones. Personally, it wouldn't take a lot to convince me that if a
>> certain set of operator names is problematic for important connectors,
>> we should avoid using those and switch to other ones. I expect others
>> on this mailing list to insist that if the connectors don't work,
>> that's the connector drivers fault for coding their connectors wrong.
>> And maybe that's the right answer, but on the other hand, maybe it's a
>> little myopic. I think the discussion is worth having.
>>
>
> In that case my vote is new operators. This has been a sore point for the
> JDBC driver
>

What if something like this was made to work?

select '{"3":5}'::jsonb operator("pg_catalog"."?") '3';

(Where the double quotes around the ? would be tolerated, which they
currently are not)

Is there a reason it can't be made to work?

Cheers,

Jeff