×
The internal search function is temporarily non-functional. The current search engine is no longer viable and we are researching alternatives.
As a stop gap measure, we are using Google's custom search engine service.
If you know of an easy to use, open source, search engine ... please contact support@midrange.com.
In response to my statement that:
>> I think we're speaking at cross-purposes here
Joe Sam Shirah wrote:
I think you're exactly right. My original "RRN is not an SQL concept"
comment was simply to note that standard SQL, and relational databases in
general, don't grok relative record numbers or even the notion of reliable
sequence without a WHERE clause.
Last night, before I had to leave for my dental appointment, I looked
through the Javadoc for java.sql.DatabaseMetaData. Interesting: it even
gives you the product name for the database system.
Actually, I don't have any *immediate* need for accessing DB2/400
through JDBC; this exercise of learning JDBC is mainly geared towards
using Scott Klement's JDBCR4 JDBC from RPG API to access "off-system"
databases.
One thing that my cursory examination of DatabaseMetaData didn't tell me
was whether there was a way (other than inferring it from the product
name, or using trial-and-error) to determine whether a database
recognizes either the MySQL "LIMIT" clause, or the similar (I forget
whose) "TOP" clause, for restricting the effect of a statement to only
the first of several indistinguishable records. (I only know of it
because that's how Sequel Pro does it with MySQL.)
As I recall, Squirrel, while it doesn't take any steps to avoid having
an update of a duplicated record affect all copies of that record, it
*does* warn you when you're about to change a bunch of duplicate records.
If a database management system truly has no access to RRNs, and
therefore cannot distinguish between to identical records, then there
should be no reason to care which one is affected by a change: "a
difference that makes no difference is no difference." If it does have
access to RRNs as a means of distinguishing two otherwise-identical
records, it should do so. The real issue is that don't want to end up
giving users with an ill-constructed database the ability to *create* a
redundancy that can only be removed by removing *all* the redundant
records and manually re-inserting them as non-redundant.
--
JHHL
midrange.com
