Just like if you give a kid a TV, the kid can find ways to learn from that for good or bad, if you give a kid a laptop, the kid can learn how to write code, how computers work, how the internet works, and eventually they'll be able to get on the internet and learn how the world works through wikipedia or whatever else happens to be on the internet. At least until they discover 4chan
If I remember correctly using O/R mapping solutions has an overhead of about 15% in response time. I have no way to sustain that number other than my memory so I will be most pleased if someone corrects me wrong.
Nevertheless, compare a 15% in performance drop with all the time saved in development by calling simple methods on an object like "persist" instead of writing whole "insert" SQL statements for a table with 30 fields.
For some kind of products that saved time (months?) is far more important than a drop of 15% in response time.
O/R mapping solutions are specially good at inserting, deleting, updating and for simple queries like "find all of this kind" or "find by this primary key"... I can tell for my whole experience that this saves tremendous amounts of time, but when you need really complex reading and association of data, like inner joins and stuff like that is when you run into trouble.
Thatâ(TM)s the reason why almost all O/R mapping solutions offer ways to use plain old SQL for complex reading queries.
The Java people even call this approach a pattern, the "JDBC for reading" pattern.
Slashdot Mirror
Yeah, EVERYBODY pays for the XP license
Google health explained by Eric Schmidt at the HIMSS conference. See for yourself.
Consider Floola, it runs pretty well on linux. (Ubuntu at least)
If I remember correctly using O/R mapping solutions has an overhead of about 15% in response time. I have no way to sustain that number other than my memory so I will be most pleased if someone corrects me wrong. Nevertheless, compare a 15% in performance drop with all the time saved in development by calling simple methods on an object like "persist" instead of writing whole "insert" SQL statements for a table with 30 fields. For some kind of products that saved time (months?) is far more important than a drop of 15% in response time. O/R mapping solutions are specially good at inserting, deleting, updating and for simple queries like "find all of this kind" or "find by this primary key"... I can tell for my whole experience that this saves tremendous amounts of time, but when you need really complex reading and association of data, like inner joins and stuff like that is when you run into trouble. Thatâ(TM)s the reason why almost all O/R mapping solutions offer ways to use plain old SQL for complex reading queries. The Java people even call this approach a pattern, the "JDBC for reading" pattern.
I just thought of the perfect soundtrack album for this game