Обсуждение: why pg_size_pretty is volatile?

On Tue, Jan 26, 2016 at 5:35 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:
> Vitaly Burovoy pointed on bug in my patch - a pg_size_bytes was VOLATILE
> function. It is copy/paste bug - I used pg_size_pretty definition, so the
> question is: why pg_size_pretty is volatile? It should be immutable too.

+1. This function relies only on the input of its argument to generate a result.
-- 
Michael

On Tue, Jan 26, 2016 at 3:00 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:
>
>
> 2016-01-26 2:00 GMT+01:00 Michael Paquier <michael.paquier@gmail.com>:
>>
>> On Tue, Jan 26, 2016 at 5:35 AM, Pavel Stehule <pavel.stehule@gmail.com>
>> wrote:
>> > Vitaly Burovoy pointed on bug in my patch - a pg_size_bytes was VOLATILE
>> > function. It is copy/paste bug - I used pg_size_pretty definition, so
>> > the
>> > question is: why pg_size_pretty is volatile? It should be immutable too.
>>
>> +1. This function relies only on the input of its argument to generate a
>> result.
>
>
> attached patch
>
> all tests passed

Pushed. Thanks!

Regards,

-- 
Fujii Masao