Re: Conversion error of floating point numbers in pl/pgsql

On Tue, Nov 17, 2015 at 9:00 PM, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
> On Mon, Nov 16, 2015 at 9:49 AM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
>> Kyotaro HORIGUCHI <horiguchi(dot)kyotaro(at)lab(dot)ntt(dot)co(dot)jp> writes:
>>> Hello. I found that 9.5 has an undocumented difference from 9.4
>>> in type cast in pl/pgsql and I think it might better be mentioned
>>> as a change of behavior in release notes.
>>
>>> Whether do you think it is worth mentioning or not in release notes?
>>
>> This seems unnecessarily alarmist to me. Anybody who's in the habit
>> of converting between float4 and float8 will already be used to this
>> behavior, because it is what has always happened everywhere else in
>> the system.
>
> Sure, but that doesn't mean nobody's functions will start behaving
> differently. It seems worth mentioning as a backward compatibility
> issue to me, because if something breaks, it may not be immediately
> obvious why it has gotten broken.

Agreed, but the note should be followed by another one warning against
any expectations of floating point behavior below the precision
threshold :-).

merlin