Linked by Thom Holwerda on Mon 4th Jan 2010 22:41 UTC
Thread beginning with comment 402538
To view parent comment, click here.
To read all comments associated with this story, please click here.
To view parent comment, click here.
To read all comments associated with this story, please click here.
RE[3]: waste of time and money
by computeruser on Tue 5th Jan 2010 12:24
in reply to "RE[2]: waste of time and money"
If your application needs to store data, it really shouldn't care how it's stored.
I used to hear this argument a lot back when MySQL had much less features than it does today, and it was wrong back then.
Most relational database systems are much more than simple data stores. And applications should care about how data is stored if they want efficient performance from a relational database.
RE[4]: waste of time and money
by google_ninja on Tue 5th Jan 2010 13:59
in reply to "RE[3]: waste of time and money"
Most relational database systems are much more than simple data stores. And applications should care about how data is stored if they want efficient performance from a relational database.
I disagree. *Many* of them are more, *most* of them are just a place to store, query, and backup data, for which an RDBMS of any kind is overkill. Hence the whole NoSQL movement.
RE[4]: waste of time and money
by rycamor on Tue 5th Jan 2010 16:49
in reply to "RE[3]: waste of time and money"
" If your application needs to store data, it really shouldn't care how it's stored.
I used to hear this argument a lot back when MySQL had much less features than it does today, and it was wrong back then.
Most relational database systems are much more than simple data stores. And applications should care about how data is stored if they want efficient performance from a relational database. "
You are right that an application should care, but performance is not really the issue. Performance tweaking of databases should be an orthogonal concern to GOOD MANAGEMENT of your data. As an example, PostgreSQL lets you save an incredible amount of application code by letting you specify constraints, complex datatypes, and logical relationships declaratively, and it has many safeguards built in to ensure that what you thought you inserted is exactly what you DID insert. Any application with mission-critical data--financial, scientific, whatever--would do well to steer clear of MySQL.
Yes, I know they have InnoDB and ANSI standard mode now, but after having been forced to work with it for months lately, I am still not impressed. As a for-instance, you cannot use the results of a MySQL stored procedure IN a subquery or view. HUH!!?? It's as if they have missed half the point of relational databases.





Member since:
2005-08-17
The application dictates the db used, not the other way around.
That sentence show exactly what is wrong with the software industry. If your application needs to store data, it really shouldn't care how it's stored.