Thread: I want more Money (the data type, of course! :-))
No, it's not spam! :-) I'm just wondering if there is a data type like Money, but with a (much much much) higher range -- checking the documentation, it would look like PG uses an int (32bits) to store the amount of cents -- but -21 million to +21 million is insufficient for accounting of a small company... :-( Float is out of the question, of course (if it is really stored as a float or double). Any suggestions? Thanks, Carlos --
Carlos Moreno wrote: > > No, it's not spam! :-) > > I'm just wondering if there is a data type like Money, but with > a (much much much) higher range -- checking the documentation, > it would look like PG uses an int (32bits) to store the amount > of cents -- but -21 million to +21 million is insufficient for > accounting of a small company... :-( > > Float is out of the question, of course (if it is really stored > as a float or double). > > Any suggestions? I normally use Numeric(10,2) HTH, Nils
Use numeric with appropriate precision information. On Sat, 19 May 2001, Carlos Moreno wrote: > > No, it's not spam! :-) > > I'm just wondering if there is a data type like Money, but with > a (much much much) higher range -- checking the documentation, > it would look like PG uses an int (32bits) to store the amount > of cents -- but -21 million to +21 million is insufficient for > accounting of a small company... :-( > > Float is out of the question, of course (if it is really stored > as a float or double).
Carlos, We use int8 and keep everything in pennies. Works for us ;) Dave ----- Original Message ----- From: "Carlos Moreno" <moreno@mochima.com> To: <pgsql-general@postgresql.org> Sent: Saturday, May 19, 2001 10:35 AM Subject: [GENERAL] I want more Money (the data type, of course! :-)) > > No, it's not spam! :-) > > I'm just wondering if there is a data type like Money, but with > a (much much much) higher range -- checking the documentation, > it would look like PG uses an int (32bits) to store the amount > of cents -- but -21 million to +21 million is insufficient for > accounting of a small company... :-( > > Float is out of the question, of course (if it is really stored > as a float or double). > > Any suggestions? > > Thanks, > > Carlos > -- > > ---------------------------(end of broadcast)--------------------------- > TIP 4: Don't 'kill -9' the postmaster >
Stephan Szabo wrote: > > Use numeric with appropriate precision information. I'm assuming that when I use numeric specifying the number of decimals, there is no rounding error in the arithmetic and storage? (well, other than rounding on the decimals beyond the ones specified -- e.g., if I say numeric, 2 decimals, there will be rounding error only in the 3rd decimal and after the 3rd decimal?) If so, then it sounds better than using an 8-byte integer to keep the pennies, given that it is more a what-you-get-is-what-you-get thing than storing the pennies, which is really a what-you-get-is-not- what-you-get-until-you-divide-it-by-100 ;-) Thanks! Carlos --
On Thu, 24 May 2001, Carlos Moreno wrote: > > Stephan Szabo wrote: > > > > Use numeric with appropriate precision information. > > I'm assuming that when I use numeric specifying the > number of decimals, there is no rounding error in > the arithmetic and storage? (well, other than That's my understanding of it. It's a fixed width exact value.
Stephan Szabo wrote: > > > I'm assuming that when I use numeric specifying the > > number of decimals, there is no rounding error in > > the arithmetic and storage? (well, other than > > That's my understanding of it. It's a fixed width > exact value. Sounds great. Thanks to all for the replies! Cheers, Carlos --