Dr. Dobbs: Questioning Traditional Data Management
About a week ago, I found in my inbox the Dr. Dobbs newsletter and since it was titled "Questioning Traditional Data Management" I decided to take a peek right away.
I have to admit, after first reading it I was a little bit upset, it felt kind of like someone walked up to me and spit in my face. But after rereading it a few times and cooling off, I believe (like most articles written) the author does have some valid points among the complaining and marketing of his book. Some of the assumptions he believes are made in "traditional data management" seem to be driven from some experience he had with some psycho data architect (and there are a few of them out there). Granted "said" psycho data architect beat these assumptions over his head, but as in most things in life you need to take a step back and think about the situation prior to just yelling fire in a crowded theater.
The assumptions he lists:
Assumption #1: It's expensive to evolve a database schema.
Assumption #2: You need to model the details up front.
Assumption #3: You need to write everything down.
Assumption #4: You need to take a data-driven approach.
Assumption #5: Review and inspections are an effective way to ensure quality.
Assumption #6: They need to govern data.
Like I said, I agree with some of his comments on these six topics but his comments tend to be more bashing the psycho data architect from his past nightmares then to most folks who enjoy the role of Data Architect.
Balance is everything in life. A good data architect along with a good application architect and project team can ensure great data quality as well as applications that perform well and meet and exceed everyone's expectations.
Until next time...Rich
I have to admit, after first reading it I was a little bit upset, it felt kind of like someone walked up to me and spit in my face. But after rereading it a few times and cooling off, I believe (like most articles written) the author does have some valid points among the complaining and marketing of his book. Some of the assumptions he believes are made in "traditional data management" seem to be driven from some experience he had with some psycho data architect (and there are a few of them out there). Granted "said" psycho data architect beat these assumptions over his head, but as in most things in life you need to take a step back and think about the situation prior to just yelling fire in a crowded theater.
The assumptions he lists:
Assumption #1: It's expensive to evolve a database schema.
Assumption #2: You need to model the details up front.
Assumption #3: You need to write everything down.
Assumption #4: You need to take a data-driven approach.
Assumption #5: Review and inspections are an effective way to ensure quality.
Assumption #6: They need to govern data.
Like I said, I agree with some of his comments on these six topics but his comments tend to be more bashing the psycho data architect from his past nightmares then to most folks who enjoy the role of Data Architect.
Balance is everything in life. A good data architect along with a good application architect and project team can ensure great data quality as well as applications that perform well and meet and exceed everyone's expectations.
Until next time...Rich
Comments