Software Engineering discussion
date
newest »


I agree with the book that persistent data means people are less likely to move away from SQL and existing databases.
I liked the page with definitions: closure, completeness, etc... That takes up a page in the book, but it was nice to get some definitions.
SQL may be one of the most expensive languages. Oracle licenses are very costly. I wonder if this is the nature of declarative languages or more inherrent of database addiction.
Page 237 made me think: Is a Google Search considered using a declaritive language?
Yes, information retrieval query languages are declarative languages for unstructured information, and SQL is an example of a declarative language for structured information. In a Google search, the user specifies what they want without any information about how to accomplish the task.
Although, the Google search language is even farther from Turing complete than SQL!
Although, the Google search language is even farther from Turing complete than SQL!
IBM System R vs. Oracle story is only briefly covered, yet it is a really classic case study on the pros and cons of corporate research openness.
This chapter did cause me to reflect on an issue Erik raised in the LUA thread. What is the definition of a computer language (both in this book and in general)? SQL (at least without extensions) is not Turing-complete (although it is Godel-complete), yet most refer to it as a data access language. I will have to do some digging for the answer.