Re: insert column monetary type ver 2

Ron Johnson <ronljohnsonjr@gmail.com>

From: Ron <ronljohnsonjr@gmail.com>
To: "pgsql-general@lists.postgresql.org" <pgsql-general@lists.postgresql.org>
Date: 2021-11-22T03:21:53Z
Lists: pgsql-general

On 11/21/21 8:43 PM, David G. Johnston wrote:
> On Sunday, November 21, 2021, Ron <ronljohnsonjr@gmail.com 
> <mailto:ronljohnsonjr@gmail.com>> wrote:
>
>     On 11/21/21 7:33 PM, Adrian Klaver wrote:
>
>         On 11/21/21 17:08, Ron wrote:
>
>             On 11/21/21 10:27 AM, Tom Lane wrote:
>
>                 Ron <ronljohnsonjr@gmail.com
>                 <mailto:ronljohnsonjr@gmail.com>> writes:
>
>
>                 The contrary argument is that people tend not to like
>                 "best guess"
>                 behavior when working with monetary amounts.
>
>
>             Best Guess is relative.  Breaking because of "ASCII Space"
>             instead of "narrow no-break space" is definitely Not Right.
>
>
>         The people I have worked for have forgiven me many sins, messing
>         up the money is not one of them.
>
>
>     Not inserting this record is also messing up the money.
>
>
> Adding processing smarts here is Not Definitely Right.  And the status quo 
> is Not Definitely Wrong.  As long as we don’t have a round-trip problem 
> I’d lean toward maintaining the status quo and accept the complaints that 
> we aren’t being as flexible as we could be with our money type.  That we 
> interpret the LC setting exactly is a defensible position to take.

Which means *that* is broken.

> Especially since money is a type we advise people avoid anyway.  We can 
> just add this to the list of reasons why.

Why then does the money type exist?  At the very least, why isn't it deprecated?

-- 
Angular momentum makes the world go 'round.