Re: Why all the max length constraints?
From: -CELKO- <jcelko212_at_earthlink.net>
Date: 29 May 2006 15:08:29 -0700
Message-ID: <1148940509.047851.130140_at_j73g2000cwa.googlegroups.com>
Date: 29 May 2006 15:08:29 -0700
Message-ID: <1148940509.047851.130140_at_j73g2000cwa.googlegroups.com>
>> Joe Celko defending the SQL types on the grounds that they were chosen to fit well with existing programming languages in which SQL would be embedded <<
It was not so much as "existing programming languages" (i.e. ANSI X3J languages, with whom we were communicating) as it was the current crop of physical data types (BCD, IEEE floating point, two-complement integers, etc.) found in the hardware, which the X3J languages and proprietary languages were using. Thus we have subtle differences in FLOAT and REAL, DECIMAL(s,p) and NUMERIC(s,p), and character set collations, etc.
If a new language comes along, it will probably be built on those hardware data types. Received on Tue May 30 2006 - 00:08:29 CEST