Hullo and welcome to this week's edition of the Log Buffer! A very special thanks to Dave Edwards at Pythian for this opportunity. Now let's begin ...

The man, the myth, the legend Tom Kyte shared some insight into a new SQL feature under consideration for Oracle: pattern matching over partitioned, ordered rows. He's asking for readers to offer up their opinion, so get to it! Pete did so on his blog as well.

Tom also just posted a lengthy write-up of some of the caveats to be aware of when using "explain plan".

Sachin, broadcasting live from his blog Oracle Online Help, has posted a nice primer on the differences between three join types that anyone who has ever look at a query plan will be familiar with: nested loops, hash joins, and sort-merge joins.

Oracle ACE Chris Foot at dbazine.com generously shared with us his top 10g tuning tools.

Kevin Closson posted a postive preview of things to come in Oracle 11g, as well as Oracle's approach.

Pete Finnigan, via his Oracle Security blog, posted two bits of worrisome information: a 0-day Oracle exploit and an Oracle rootkit are out and about.

Those of you interested in Oracle client-side load balancing might want to head over to yas' Oracle Today.

Robert Vollman at oracleblog snuck in a few Oracle beefs, and also gives a positive mention of this blog! Huzzah!

Coskan Gundogar, from Istanbul (not Constantinople), has some questions for new graduates that think they'd like to be DBAs.

Switching away from Oracle to MySQL, Brian "Krow" Aker posted some hot news about a Postgres storage engine for MySQL. It seemed an open-source enthusiasts utopia, until everyone realized what day it was (although I was ready to buy one of these).

The folks at HackMySQL share with us a script for checking databases or tables for unused indexes: mysqlidxchk.

The aforementioned Tom Kyte even shared a recent MySQL experience with us, and Pythian's Paul Vallee shared a similar experience. Neither bode well for the MySQL fans in the audience.

Zack Urlocker at TheOpenForce.com writes that Paul Tuckfield from YouTube will be giving a talk at an upcoming MySQL conference on how YouTube scales with MySQL.

Jeremy Cole opened our eyes to a more efficient alternative to MySQL's SHOW VARIABLES.

Fans, or even casual acquaintances, of MySQL replication will find joy in Peter Zaitsev's coverage of the MySQL Master-Master Replication Manager. You'll also definitely want to help out MySQL's own Robin Schumacher and give him some feedback on replication monitoring.

Our PostgreSQL fans will have surely seen Robert "xzilla" Treat's tip at PlanetPostgreSQL about monthly log rotation with PostgreSQL's built-in logging.

Both "The SQL Doctor" and Kimberly Tripp posted about Microsoft SQL Server 2005's online book search.

Decipher Information Systems' blog had an interesting post contrasting MS SQL Server and Oracle with regard to computed columns.

Craig Mullins, host of the DB2 Portal Blog, wrote about a couple of tricks and treats in DB2. First up was mimicking ROWNUM in DB2, followed later in the week by a treatise on INTERSECT and EXCEPT in DB2. But he didn't stop there, and returning readers were rewarded with a review of ORDER BY and FETCH FIRST in DB2 subselects. And just when I thought I was out, he pulls me back in with a write-up on DB2's native XML support!

Over at his ITtoolbox blog, Willie Favero asks a question that probably shouldn't be asked.

And I think we'll end this edition of the Log Buffer right there. Probably fitting that I host the April Fool's week issue! I do apologize for the lack of PostgreSQL coverage, not a lot came across my desk this week. It only serves to remind me to make some time to play with PostgreSQL!

I hope you've enjoyed your stay at my meager blog and will return again to read about my many misadventures as I stumble into more Oracle parameters that I've let go untuned for years! For those of you who have made it this far, I'll reward you with some footage of my son's horse impression. Have a great weekend!