You have probably come to this site because you are considering buying Oracle for your company. Most likely, an employee of yours has asked you to reconsider. Most likely, you have already made up your mind. You are going to buy Oracle. Most likely your reasons are these: Oracle is what the whole world uses. It is big. It is rock solid. It is scalable. It will handle everything the company needs. Oracle will be around forever.

All your reasons are true . . . but not in the way you think. Oracle is what the whole world uses, but not because of Oracle's technical merits. The Oracle product is big, but not because of robust functionality. Oracle is rock solid, but not because the software is solid. Oracle is scalable, but not because of good design. Oracle handles everything your company needs, but Oracle handles everything less effiectively than its competitors.

I am a CTO. In the end, the most important thing I can tell you is, is that Oracle is a money pit. I ask you to not commit yourself to Oracle. If you commit yourself to Oracle, know that you do so for marketing purposes only. I have nothing personally against Oracle. I just have an indignation towards poorly designed software that induces needless and costly trouble.

I have listed many technical problems on these web pages. Each problem will cost your computer money. Some problems will cost you up to several US$ 100,000. There is no doubt of this. You must understand that each problem comes only with Oracle. Other databases like Sybase or Microsoft do not have these problems. If you are skeptical at this moment , then in the back of your mind you may be thinking that Sybase or Microsoft has made some other mistake or limitation which offset Oracle's. But that thought iwould be wrong. Only Oracle has the deficiencies.

As a CTO, you may not have time to read these web pages. I'll summarize:

1. The whole world uses Oracle because 1) people want to join the biggest band in town regardless of how the band sounds, 2) Oracle requires an army of developers and DBAs with specializations to deal with the irregularities of the software, and 3) there are a whole lot of irregularites. I described a full circle. The reason why there are so many Oracle professionals, that the Oracle newsgroups are full and that Oracle books are plentiful, is not because people want Oracle, but rather because people require help.

2. The real reason why Oracle software is big, is because 70% of the software is Band-Aids to cover over the 30% that doesn't work. The reference manuals reveal a cornucopia of commands and options. But 70% of these commands and options exist as work-arounds. In turn, the work-arounds only work some of the time. So the inconsistencies require explaining . . . and more commands. And so Oracles is big.

3. The rock solidness of Oracle depends on how you look at Oracle. From a business perspective, Oracle is a rock solid company. But it stops there. Don't confuse the company with the product. The Oracle software itself has been replastered and repatched many times over. The Marketing Department does a good job painting. So while the Oracle software gives the appearance of solidity, it takes a professional about one hour to realize that the solidity is skin deep. There is no foundation.

4. The real reason behind Oracle's scaleability is because it lacks it to begin with. Oracle's ability to handle work starts out magnitudes slower than its competitors. For example, it takes 180,000 Oracle processes to handle 32,767 databases (at least 6 processes per database). As for the competitors, it takes a single process for Microsoft or Sybase to handle all 32,767 databases. That's right. 1 Sybase ASE handles 32,767 databases. Since a single computer cannot really handle 180,000 processes, Oracle gives you the scaleability to spread the work out over several computers. Oracle has to spread out. Sybase and Microsoft can also spread the work. They just don't need to.

5. It is true that Oracle can handle everything that your company needs. It will take your programmers three times longer to figure out the solution. The solution will run three times slower. You will need three times the staff to maintain the solution.

To Developers:

To you who are SQL Server and Sybase programmers, I take it your CIO wants to switch to Oracle. You have started playing with Oracle. And now you cannot believe that Oracle, the most popular database in the world, lacks even the simpliest things. You cannot believe the time it takes you to program basic tasks. You cannot believe that you must code elaborate time-consuming inefficient work-arounds. You cannot believe that common sense functionality taken for granted in Sybase and SQL Server, does not exist in Oracle.

Your suspicions are true. Oracle is the worst commericial database in the world. Though it is the most popular, Oracle is the most technically inferior. It indeed lacks real-time print statements. It lacks batches. It lacks a single SQL language. In place of a single language like T-SQL, Oracle does have three separate languages with syntax inconsistent with the other. And what idiot Oracle designer redefined "null as NOT the empty set? Did they think they could just change the laws of mathematics and get away with it? With that misdefinition, one cannot even model a simple phone book. (Think . . . Columns in a primary key must not be null. The primary key of a phone book is Last Name, FIrst Name, Middle Initial. Not everyone has a Middle Initial. According to Oracle, an empty string is null. Whoops! )

These web pages provide you, the developer, with about 50 specific Oracle problems. It is best to get familiar with them before you start a project. If you know about these problems up front, you will avoid pitfalls later. You will save man years of development .