Re: pg_dump and psql disagree on numerical precision (powerpc) - Mailing list pgsql-bugs

From Christof Petig
Subject Re: pg_dump and psql disagree on numerical precision (powerpc)
Date
Msg-id 3FB4F8FE.6020209@petig-baender.de
Whole thread Raw
In response to Re: pg_dump and psql disagree on numerical precision (powerpc)  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: pg_dump and psql disagree on numerical precision (powerpc)  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-bugs
Tom Lane schrieb:
> Christof Petig <christof@petig-baender.de> writes:
>
>>COPY test (x) FROM stdin;
>>0.29999999999999999
>>\.
>
>
> This is not a bug; it's pg_dump trying to recreate the actual floating
> point value as accurately as it can.  See the discussion of
> extra_float_digits from six months or a year back.

Oh, I see (and generally appreciate). Thank you for reminding me.

Commenting out "set extra_float_digits=2" in pg_dump.c gives me the old
behaviour. Is it only me wishing for a command line parameter for
pg_dump to set desired precision?

    Christof

pgsql-bugs by date:

Previous
From: Tom Lane
Date:
Subject: Re: pg_dump and psql disagree on numerical precision (powerpc)
Next
From: Tom Lane
Date:
Subject: Re: pg_dump and psql disagree on numerical precision (powerpc)