Michael Sansoterra
Michael Sansoterra is a DBA for Broadway Systems in Grand Rapids, Michigan.
-
More V5R3 SQL Enhancements
July 14, 2004 Michael Sansoterra
My last article briefly covered the new Binary and Varbinary data types, some new scalar functions, and the sequence object that is new to the iSeries implementation of DB2. The good news is there are more SQL enhancements to cover!
DISCLAIMER
Once again, since I don’t yet have access to a machine with V5R3, this information was gleaned from the new V5R3 manuals and may be subject to revision! I’d love to hear from someone who has a beta of V5R3 who can offer more insight on these features.
NAMED COLUMN JOIN
A new shorthand join syntax called a “named
-
V5R3 SQL Enhancements
June 30, 2004 Michael Sansoterra
Like a kid at Christmas, with every release of OS/400 (now i5/OS), I’m always eager to find out what enhancements IBM has made to SQL. Once again, IBM has been very good to SQL programmers with this latest release. Here is a review of many of the latest enhancements to SQL.
DISCLAIMER
Since I don’t yet have access to a machine with V5R3, this information was gleaned from the new V5R3 manuals and may be subject to revision if I interpreted something incorrectly! Where possible, I tried to validate the syntax of the material by testing some of these features
-
Performing Numeric Bitwise Comparisons in SQL and RPG
April 14, 2004 Michael Sansoterra
[The code for this article is available for download.]
In the iSeries world, there isn’t a high demand for propeller-heads to write software that does bit manipulation. I’ve used bit manipulation on the iSeries a handful of times–once while writing a BASE64 routine to create binary e-mail attachments, once while communicating with a shop-floor device, and a few of other times for doing things like fixing decimal data errors.
Each attempt to work with bitwise operations in RPG or SQL resulted in a new frustration. That’s because both environments do their bit manipulation and comparisons against strings,
-
Delimited List Processing in SQL
March 24, 2004 Michael Sansoterra
[The code for this article is available for download.]
I recently worked on the database side of a Web project for which the Web developers needed to pass several item numbers to an iSeries for processing. Traditional methods would have dictated these items be inserted into a temporary or transaction file, then converted and processed as an XML file or passed individually to a stored procedure for processing.
An alternative that is fast and painless (provided the related item information is minimal) is to make a single call to a stored procedure and to pass the item
-
Practical Use of a Translation Table
March 3, 2004 Michael Sansoterra
While working on a Microsoft Access 97 project, I was purging and redownloading a large amount of data every night from an iSeries. I noticed that once the data was downloaded, Access took a while to query the data even though there were proper indexes built on the table.
Some investigation revealed that Access was actually doing a large amount of re-sorting of the data because of the difference in the EBCDIC and ASCII collating sequences. In particular, EBCDIC places numerals at the end of the sort order; whereas, ASCII places them toward the beginning. So Access still had to