NoCOUG

I attended the Northern California Oracle User Group (NoCOUG) conference last Thursday, in Mountain View. The event took place at the Computer History Museum (more information here) and was informative.

The museum in itself was well worth the commute (Oakland to Mountain View round trip is about 70 miles) as they have an amazing collection of "old" computers and a well documented presentation of the challenge the building of a strong chess program on a computer represented. If you have a chance to be in Mountain View some day, I strongly encourage you to take the time to have a tour. Of course, this only applies if you are a computer lover like me otherwise, you may not be that thrilled by the visit.

I attended two presentations by Tom Kyte, one of the most knowledgeable individuals in the Oracle field. He is the Tom behind the site asktom.oracle.com.

The first presentation he made was about asking "why?" when someone when someone asks him how to do X or Y with Oracle. I guess his intention is to protect users from themselves by getting as much information as possible up front.

I enjoyed his presentation very much because I usually have the same approach when someone asks me a question that starts with "How do I ..." or, even better, when someone tells me "we need to do this and that" without any further explanation.

Immediately, I try to get more data, find out where this person is coming from, what thought process was followed to arrive to the conclusion that is summed up in the question I am being asked. I usually ask "what is your objective". Many people get upset pretty quickly, as they only want an answer and think that I am trying to avoid the question. The whole point of asking about the true objective is definitely not to avoid answering but to define the context.

It may be due to my education as an engineer, but to me, pushing a solution before analyzing the problem is just ideology.

Tom's second presentation was about binding when building an application on top of an Oracle database. Once again, it is all a question of putting things in perspective. It is just plain amazing to see how many people don't take the time to learn the minimum needed to use Oracle (or any other sophisticated system for that matter) before starting their project.

I know that there are many constraints to get the system up and running as soon as possible (and for a good reason, time is money) but come on! If you don't take the time to learn how to do your job correctly, your system will never be up and running and you will end up playing the infamous blaming game (pick a scapegoat: Oracle, the Operating System, the disks, the CPU, the business users, management etc.) when the project was doomed to fail from day 1.

I have been using computers for a long time now, as I had my first machine in 1985 when I was still a kid. Compared to the machines we can buy nowadays, this one looks like a toy! Anyway, the point here is that I learn early on how to get the maximum out of a machine and that one should always be as thrifty as possible when building a system. I hate it when I hear comments like "the server is very powerful so we don't need to...” Sure, the server is powerful and your stuff runs great when you are alone using. Now, how does it behave when you have 200+ users pounding it?

I am currently reading (among other books as I usually read more than one book at a time) the latest book by Tom Kyte. My current interest is about locking mechanisms, read consistency and so on. Another interesting area where a lot of people never spent the time to understand how Oracle is working and what the consequences are. Bottom line: You get ridiculous myths that appear here and there about what Oracle is doing or not behind the scenes.