From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | mike(dot)griffin(at)mygenerationsoftware(dot)com |
Cc: | "Bruno Wolff III" <bruno(at)wolff(dot)to>, pgsql-general(at)postgresql(dot)org |
Subject: | Re: 7.4's INFORMATION_SCHEMA.Columns View |
Date: | 2004-06-19 01:30:13 |
Message-ID: | 5725.1087608613@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
>>> The SQL spec doesn't allow unconstrained lengths for these types
>>> so it gives no guidance about what to display in the information_schema
>>> views. Any opinions?
>>
>> If there isn't a set scale for the type, then NULL would probably make the
>> most sense.
After more thought I like returning NULL for both precision and scale in
the case of unconstrained numeric columns. Any other value is
arbitrary. In particular, the 1000 cited in the docs is *very*
arbitrary, and I don't think it actually constrains what you can store,
only what you can declare as a column precision. [tries it...] Yup,
I can store "power(10.0, 10000)" in an unconstrained numeric column.
It seems to fail around 10^140000 but I'm not sure where that limit
is coming from exactly...
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2004-06-19 01:45:38 | Re: RHEL 2.1 rpms for 7.4.2 |
Previous Message | Tom Lane | 2004-06-19 01:18:25 | Re: 7.4's INFORMATION_SCHEMA.Columns View |