Following up on a post from last week, our resident SQL guru Rudy Limeback provides some more thoughts on the matter of what’s so bad (or so good) about SQL:

Yes, I ♥ SQL, primarily because knowledge of this ubiquitous language allows me to be immediately productive in working with so many different database systems.

Imagine if each time you wanted to obtain information from a different database system, you had to learn a completely different language.

Instead, all that is necessary is to learn the small, inconsequential differences between one database system’s implementation of SQL and the next.

MySQL  
T-SQL  
UDB SQL  
PostgreSQL  
PL/SQL

This is due in no small part to the effort of many people and vendors to define the SQL standard. That so few database systems faithfully support all aspects of the standard is no serious knock against the advantages of having a standard language.

Further, SQL is drop dead simple. There are only a few basic clauses:

SELECT

FROM

WHERE

GROUP BY

HAVING

ORDER BY

As well, you’ll need some simple concepts like INNER and OUTER JOINs, UNIONs, and subqueries. With these basic tools, you can solve even the most
complex data retrieval problems.

That said, it’s a lot like chess. You can learn the basic moves in half an hour, but you need to invest the necessary time to improve your skills. Practice, practice, practice.

It looks like we’re making some progress toward debunking the “SQL sucks” myth!

-Elisa Gabbert

If you read the mainstream media, the Daylight Savings Time bug will lead to the apocalypse that Y2K never was.

Hype, obviously, but nevertheless DST is a serious problem that all Oracle shops need to address. Besides the operating system problems, architect Chris Foote points out what the DST bug affects in Oracle:

•  JVM
• Time zone files
• Grid Control agents
• E-Business Suite
• Application Server

Unfortunately, the quantity of rants in the blogosphere indicates that Oracle’s response to the DST problem has been muddled at best. For example, over at the Yet Another Oracle DBA blog, Herod writes:

Is it just me, or is Oracle’s effort towards the DST seem to be a convoluted mess of notes, readme’s, more notes and superseding items. Download this, this, this, this, and this, and possibly this but only if you have this and this. 

Sounds like fun, especially if you have dozens or hundreds of systems to patch, like this blogger, who writes:

As someone who was forced to write DST checking scripts for an Enterprise of 1500+ Oracle databases, let me just say how utterly weak it is that Oracle has identified a DST 2007 issues with their database and provided barely any tools to help sniff around for it. What have they provided? A script to tell you if you have affected rows… A couple of other rag tag scripts here and there if you dig hard enough…. Pathetic.

If you’re just tackling the problem now, you’re up the proverbial creek (Mike R. has a good list of Metalink resources to get you started). But for those who have already completed the task, how did it go?  Is it a non-issue or an emergency? Any tips or tools that are essential?

Cheers, Tim

The other day I stumbled across a blog post titled “In defense of SQL” — in this post, blogger Kenneth Downs asks, “So am I the only one who likes SQL?”

Is he? Apparently SQL has a bad rap among development languages — people tolerate it because they figure they have to, but they think it, well, sucks. Downs, however, lists a number of things to enjoy about SQL. For instance, it can be easily generated with other languages, and it’s easy to read and write.

I asked a couple of our site experts how they feel about the structured query language. Like Downs, Brian Peasland, DBA and expert in database design as well as backup and recovery, also has a soft spot in his heart for SQL:

I have always been a fan of SQL. When I was studying database theory, I was taught that there are three main operations in any database query: 1) selection, 2) projection and 3) joins. Think of a table as a grid with each record as a row in the grid and each attribute as a column in the grid. A “selection” is a vertical slicing of the grid. You simply indicate which columns you want. Not surprisingly, the selection is the entire purpose behind the SELECT clause of a typical SQL statement. A “projection” is a horizontal slicing of that grid. The projection defines which records you returned out of all possible records in the table. The projection is defined in the WHERE clause of the SQL statement. Many who know SQL already know that a “join” operation is taking more than one table and stitching them together. The FROM clause defines the join operation … or the table(s) to find the data. So at its basic level, a simple SQL statement looks like the following:

SELECT the_columns_you_want_to_see
FROM the_tables_that_contain_the_data
WHERE some_conditions_are_true;

Once you know the basic tenet above, reading and writing SQL statements becomes much easier. The FROM clause tells you where to pull the data *from* and the WHERE clause tells you which rows to pull out of your data source. Once you know the rows of data to be returned, we only show what the user SELECTed. Once I learned the basics of SQL and started thinking of SQL similar to the above, it really made things easy for me. And it just makes sense. I agree that one can get some convoluted SQL statements which do require some analysis before truly understanding what is going on, but most SQL statements are not that complicated.

One other thing that I think is great with SQL is its standardization. I have worked in Oracle for a very long time. When I first started working with SQL Server, I did not have to learn an entirely new query language. This was huge! Granted, not all SQL statements work in both systems, but many do. Some things are different, but I just had to learn a few things to easily rewrite my query. For instance, the substring function in Oracle is SUBSTR, while in SQL Server, it is SUBSTRING. In Oracle, you use INSTR, but the similar function in SQL Server is CHARINDEX. So a few non-standardized functions are different. But for the most part, my SQL queries work well in both.

I am well aware of the ANSI SQL standards and that each RDBMS vendor does not implement all of the standards, so not all SQL statements work the same in every database. But the point is that even if the SQL statement works in SQL Server but not in Oracle, I do not have to redo everything to get that SQL statement to work in Oracle. I just have to tweak it a little to get the statement to execute.

Finally, most of the people that I have worked with that do not like SQL are those who did not really get a chance to learn it. They were taught the simplest of SQL statements and then let loose to code their application. As their application became more complex, coding a SQL statement became tougher, and more frustrating. But those I have seen take a 1-week course on SQL for their RDBMS of choice do not have this same attitude. I think this speaks volumes about why some hate SQL and others like it.

All you other DBAs and developers out there: what do you think? Is SQL not so bad or does it deserve its nasty reputation?

(Stay tuned for remarks from SQL guru Rudy Limeback on this topic.)

-Elisa

The first round of Oracle Fusion applications will enter the market in 2008. Is your company ready?

According to Oracle executives, industry analysts and several published reports and blogs, there are numerous things to consider as companies decide whether or not to upgrade to Fusion — Oracle’s plan to overhaul all of its existing and newly acquired applications to run on Oracle Fusion Middleware.

Reasons to upgrade

If your firm is committed to services oriented architecture (SOA), then experts say Oracle Fusion is the way to go. In fact that’s the main goal of the redesign — to give Oracle customers a Java based, modular platform upon which to build an SOA.

Experts say another benefit of Fusion is that companies will be getting the “best of the best” features and functionality that Oracle has acquired in recent years. That means the best of PeopleSoft, Siebel, JD Edwards and about 20 other acquired companies.

Another reason to be keen on Fusion is that the host of newly designed applications will incorporate business intelligence and other analytical software giving users a better understanding of their respective businesses.

Reasons to wait

On the other hand, as with most new software releases, many companies will opt to hold off on Fusion while the early adopters iron out the kinks.

Also, Oracle has promised unlimited support for all of its acquired applications regardless of whether or not people upgrade to Fusion. Although some doubt that Oracle can keep innovating these standalone apps forever, Oracle says it will be no problem. If your company only really needs one of these apps, then maybe Fusion isn’t worth it to you.

Finally, those in the know point out that Oracle Fusion experts will be few and far between and expensive to employ. It may also be costly to train existing employees on how to effectively manage and make use of Fusion applications.

What say you?

Is your company steadfast in it’s march toward Fusion, or does it plan to play a game of wait-and-see? I’d like to hear your opinions on this topic for a feature story I’m writing. Tell me about your Fusion plans and concerns and I’ll talk to some experts to get a response. And remember, ready or not, Fusion is coming.

– Mark Brunelli, News Editor

To paraphrase the Chinese aphorism, “Be careful what you mock, it just might come true.”

A few weeks ago, I declined to invent a “soon-to-be-hackneyed phrase like BI 2.0.”  How clever! Well, it seems that multiple pundits and analysts had already beaten me to it. Here is a sampling:

• BI 2.0: Pervasive, intelligent and timely
• BI 2.0: The next generation
• Business intelligence 2.0 and enterprise decision management
• In search of insight
• Business intelligence 2.0: Simpler, more accessible, inevitable
• Gartner: It’s business intelligence 2.0 time

So what is BI 2.0 anyway? Here’s one of several definitions:

The current [BI] offerings are based on a personal, desktop-based metaphor of a one-to-one relationship between an analyst and retrospective view of information with a hoped-for direct link to decisive action, but such instances are rare. BI 2.0 is positioned to be proactive, real-time, operational and integrated with business processes that can extend beyond the firewall as easily as providing simple, personal analytical tools on an as needed basis with a minimal footprint and cost.

As with all trendy technologies, the goals of BI 2.0 are laudable but  easier said than done (much easier!). Nonetheless, “proactive, real-time, operational and integrated” sounds like music to the ears of many a user, so this buzzword might actually have some legs.  I’ll try to refrain from mocking it and give it a chance… for now.

Cheers, Tim 2.0

Oracle and other major software vendors like SAP, IBM and Microsoft are either making or planning to make a big deal about Web 2.0. And, according to one prominent industry analyst, a primary motivation for this may be their desire to stave off an impending attack from search engine giant Google.

Google, says Jim Murphy, an analyst with Boston, Mass.-based AMR Research, has started putting together a suite of collaborative enterprise applications. And while no one knows exactly what the suite will look like, Murphy says it’s a safe bet that it will incorporate so called Web 2.0 technologies like blogs, wikis, and RSS feeds.

“Google is poised to introduce an alternative to the traditional way of getting at enterprise information,” Murphy said. “They’ve been demonstrating this capability on the external consumer Web for some time and, if you’re talking about Web 2.0 types of trends, [some] of the expectations really come from the Google experience.”

Oracle this week introduced its new WebCenter Suite, a set of tools that adds an interaction layer to Oracle applications, giving users the ability to collaborate via blogs, wikis and RSS feeds. Oracle says WebCenter will serve as the foundation of the user interface to Oracle’s Fusion Applications when they debut beginning in 2008.

And while providing insight into the user interface is a significant step forward on the road to Fusion, says Murphy, it’s not Oracle’s only motivation.

“They’re all trying to fend off Google,” he said.

What do you think? Does Oracle have something to fear from Google? What about IBM and SAP, or Microsoft for that matter?

– Mark

Great news for Oracle users, newbies and otherwise, everywhere: the official Oracle documentation is currently being indexed by Google. You used to have to register for the Oracle Technology Network (OTN) to gain access to these docs. Registration was free, but still an annoying extra step. Then Oracle introduced “Google-style” searching, but the docs were still unavailable to external search. Soon we’ll all be a mere Google search away from this wealth of information. Three cheers! (Unfortunately, word is that MetaLink will never be similarly indexed.)

-Elisa

So Oracle thinks “wikis” and “mash-ups” are going to storm the enterprise? That was some darn quick bandwagon jumping — those terms have barely entered the popular lexicon!

In case you hadn’t heard, Oracle’s new J2EE-based WebCenter Suite is the newest component of the Fusion Middleware platform and is available as an option for the Oracle App Server. Oracle says the suite lets developers create an environment that increases worker productivity by combining Web 2.0 services with transactional processes, business intelligence data, communications such as instant messaging and RSS feeds, and structured or unstructured content. Reaction to the release has been mixed: positive and negative.

Sorry to rain on Oracle’s parade, but these sorts of “Web 2.0” technologies may be great for waxing eloquent about one’s cat, or about the latest band on the teen goth scene, but will they ever become a useful presence in the corporate world. And should they?

The notion of “Enterprise 2.0″ has been floated out there in the blogosphere, defined as:

…the use of emergent social software platforms within companies, or between companies and their partners or customers.

Some examples include:

• DrKW’s internal blogs and wikis
• Rite Solutions’ prediction markets
• Enterprise tagging
• R&D departments’ use of  Innocentive to find solutions to problems that have been stumping them.
• MK Taxi’s ability to connect mobile phone users in Tokyo directly to the driver of the cab closest to them, bypassing the dispatch center altogether.
• Employee blogs

However, so far it seems like a stretch that blogging and wikis will become more than just a trendy diversion and productivity-killer. Given the extreme time crunch faced by the vast majority in corporate IT these days, who has the time to do this? Do you?

Of course, I’m ”old school” and could very well be wrong!

Cheers, Tim

In an earlier post covering recent Oracle rumors, I linked to one blogger who believed a MySQL version of Unbreakable Linux could be a good thing for everyone (not just Oracle). Now another voice, Brian Profitt from ServerWatch.com, has chimed in with a similar opinion.

Profitt doesn’t pull any punches — he makes it clear that he believes Oracle’s reasons for offering support for open source products are less than altruistic (his “initial reaction,” he says, was “Oracle = Arrogant”). Nonetheless, he argues, these “Unbreakable” offerings are ultimately beneficial for the market and the enterprise — Oracle is bringing us closer to the dream of “one-stop shopping,” a world where customers can go to a single vendor for all their products and support. The open sourcers “should have thought of doing this before,” he says. But they were too busy playing by “the old rules.”

Mike Olson, Oracle’s VP of embedded technologies, responds to this article on his own blog. Surprisingly, he doesn’t jump to Oracle’s defense or refute Profitt’s claims. He writes:

Brian talks about selfish and altruistic motivation, and asserts that Oracle acts selfishly. He’s right, and it’s important to understand why.

Olson contends that all public companies invest in research and development for the same reason: to drive profits and shareholder value. However, the effect of this is, he claims, “indistinguishable from altruism.” It’s in Oracle’s interest to keep its customers happy, so Oracle delivers better products and service. If everyone reaps the benefits of this, who cares if the motives are selfish or not?

(Olson here focuses on Oracle’s support plan for Linux; he doesn’t affirm or deny the rumors about MySQL.)

What do you think? Do selfish corporations contribute to a healthier, happier market? Or is there still something to be said for the old rules and fair play?

–Elisa Gabbert, Assistant Editor

Although the usual suspects continued to dominate the new 2007 Gartner Magic Quadrant for business intelligence platforms, changes are afoot that may shake up the status quo.

Several trends seem to be conspiring to re-write the BI vendor landscape:

• The convergence of BI and corporate performance management
• The convergence of BI and content management
• The convergence of enterprise search and BI
• The convergence of BI, EII and SOA
• The decline of the pure-play BI vendor

Notice a pattern here? Is there a “next-generation” BI emerging that is much broader in scope?  (No, I’m not going to invent a soon-to-be-hackneyed phrase like “BI 2.0″!) As the industry continues to move beyond basic reports and metrics toward process- and strategy-driven BI, can the pure-play vendors survive competing against the large platform vendors such as Oracle, Microsoft and SAP?

On the one hand, of course they can — if they can adapt and innovate faster and release higher quality products. Or they could just buy innovation, as Cognos’ purchase of the BI appliance vendor Celequest recently showed.  On the other hand, Oracle could then just buy Cognos…

Given the huge spending in BI systems now, the stakes are being raised and the outcome is uncertain. However, if BI continues to morph into the essential cog in an integrated enterprise architecture machine — as Oracle is driving at with their recently-announced Oracle Data Integrator (ODI) – then the large platform vendors have the clear advantage.

Cheers, Tim