Thursday, February 14, 2008

Love/hate, Part I

I have this love/hate relationship with Oracle. Although I rarely hear it, I would find it hard to believe that anyone who works with Oracle for very long does not feel the same way in these days. But hey, perhaps I am wrong and the people who gush over Oracle really do just have a passion for the software.

Allow me to ramble a little. This is a blog after all.

On the plus side, I have learned a lot, and that Oracle RDBMS-specific education has really boosted my career. While I still have a long way to go, I have come a long way from the Sybase newbie I started out as, not having any clue what SQL meant, let alone what it was or how to use it. I gained a solid appreciation not only for all the hard work that goes into all the cool stuff that Oracle does, but the wealth of support from the User community as well. There are some really slick, intelligent and brilliant professionals out there who are not Oracle employees. And not merely a handful, but a throng.

Now a negative. The folks who are leading Oracle are way too ambitious in my opinion. They want to control everything. There is way too much emphasis on the bleeding edge, and not enough support for the "tried and true". I am a DBA, and my niche focus hardly ever veers off into other Oracle products like BI, AS, SOA or any other slew of acronyms. If the database server is truly Oracle's flagship product, it is not because of its superior quality, in my opinion, but rather because they charge an arm and a leg for support. I could write a whole book about Oracle Support.

Another positive thing. Oracle's RDBMS is fodder for some most excellent theoretical and scientific arguments. Watch some interchanges between Oracle gurus on the usenet, oracle-l or the Oracle forums; some of those discussions are worthy of a doctoral thesis. Oracle does some really mind-blowing things, especially in the areas of manipulating data. In general, the software is fast, scalable and rather solid.

And the pendulum swings. Oracle tries to do too much. Let's look at the RDBMS. 8i came out in 1999, 9i in 2001, 10g in 2003, in 2005, 11g in 2007. Each new release introduced hundreds of new features, and thousands of new bugs. The worst part, every two years a stable-but-older version is falling off the support list. This problem is exacerbated within patch levels, where the rate of turn-over seems to be about 1 patch per year, each with new features and new bugs, killing support for the new two-off patches. Oh how I wish Oracle would commit to sticking with a "stable" version, making it bullet-proof.

All these new features are a double-edged sword. The media loves "new features", Oracle Support, I am sure, hates them. As a customer, user and DBA, I love the new technologies and the learning opportunities, but I hate how there is often poor documentation, lack of support and a host of accompanying bugs.

Am I complaining? You bet. Am I a whiner? I could care less. Taking a step back, I have a big beef with software development in general. Example, Microsoft Vista. Like the Grammy's, there should be awards for best-written and most-stable products. Or perhaps that would be too depressing.

In my opinion, one of the most ironic statements to come out of Ellison's mouth was the whole "War on Complexity". Oracle definitely lost that war. Look at the monstrosity called "Oracle Enterprise Manager". Sure, its free, and that is why other companies are capitalizing and making money by offering a commercial, better product for managing databases. If there really is a war on complexity, then I will walk to the moon. What I see is a war on simplicity. And it is not even fair to call it a war; rather, a wholesale rape.

I have had the opportunity to interact directly with Oracle employees at various levels. For the most part, I have been blown away by the graciousness, intelligence and customer-focus these men and women have. In talking with them, I have been convinced that that there are a lot of good people working hard to make an excellent product. What I dislike is the general direction of the product, and the decisions being made that influence it thus. And again, for the record, when I say "the product", I only really care about the RDBMS. I do not have anything good to say about other offerings. =)


Anonymous said...

Totally agree.

I want to use Oracle features to solve problems. Not to spend my time solving problems with Oracle features (Bugs).

Oracle need to concentrate on stability, not features for a while.

Andy Campbell

Charles Schultz said...

*grin* Or, at least maintain support for terminal releases. Obviously, that is a paradox because if they support a "terminal release" and apply bug fixes, Oracle will most likely release a new patch level for that version.

Is it just me, or are there many similarities between Microsoft and Oracle?