On 1/24/19 7:48 AM, Geoff Winkless wrote:
> On Thu, 24 Jan 2019 at 15:40, Adrian Klaver <adrian.klaver@aklaver.com> wrote:
>> delete from delete_test where
>>
>> and then forget the 'field =' part. Though my more common mistake along
>> that line is:
>>
>> delete from delete_test;
>>
>> At any rate, if it can be done it will be done.
>
> If you follow that logic, then having a single boolean test at all
> should be invalid.
>
> CREATE TABLE mytest (myval char (1));
> INSERT INTO mytest VALUES ('a'),('b'),('c'),('s'),('t');
> DELETE FROM mytest WHERE 't';
> SELECT * FROM mytest;
> myval
> -------
> (0 rows)
People are going to make mistakes that is a given. Eliminating a boolean
test is not going to change that. Where this particular sub-thread
started was with this from a previous post of yours:
"My own opinion is that non-0 should implicitly cast as true and 0
should cast as false. ..."
That opens an infinite number of values that could be seen as True. That
in turn leads to greater chance of fat-thumbing yourself into an oops.
Like you say it is a matter of opinion. The projects opinion is here:
https://www.postgresql.org/docs/11/datatype-boolean.html
and it works for me.
>
> Geoff
>
--
Adrian Klaver
adrian.klaver@aklaver.com