Re: Relational Inheritance Thoughts
От | Richard Huxton |
---|---|
Тема | Re: Relational Inheritance Thoughts |
Дата | |
Msg-id | 43BEAFCF.7060603@archonet.com обсуждение исходный текст |
Ответ на | Relational Inheritance Thoughts (Trent Shipley <tshipley@deru.com>) |
Список | pgsql-general |
Trent Shipley wrote: > At the risk of belaboring the point, relational single inheritance resembles > a hierarchical database, like a traditional computer file system. A > relational multiple inheritance database resembles a network database, like > the once promising CODASYL standard or the GROVE or DOM representation of an > XML hyper-document. > A database with relational multiple inheritance is a hybrid between a > relational database and network database that ought to support any mixture of > the data models without prejudice. I'm not sure any of this is true. Relational inheritance (where one table "inherits" the attributes of another) is a syntactic shorthand, nothing more. You could re-implement it either with one "master" table and several "detail" tables or unions of several tables with shared attribute-names. In a network-database or document, such as an XML document you use the hierarchy / network edges to navigate to individual data items, not to "navigate" (and I don't think the term is correct here) your schema definition. -- Richard Huxton Archonet Ltd
В списке pgsql-general по дате отправления: