On Mon, Feb 28, 2005 at 05:05:37PM -0000, Tim Anderson wrote: > SELECT Name, Title, Books.ID, Authors.ID FROM Books inner join Authors > on Books.AuthorID = Authors.ID ORDER BY Authors.Name, Books.Title; > > In this case, the query is unambiguous, but by default Sqlite returns > the column names as: > > Name > Title > ID > ID
Well, that looks like correct behavior to me. If you wanted the column to be called something different you should have done "... Authors.ID as Authors_ID" or something like that in your query. So if these various db interface layers need special features from SQLite in order to rewrite those column names, then it must be because the users of those interface layers are writing bad queries, queries that fail to specify the unique column names that the users actually needs or wants. Why is that? Having the user application correctly tack on an "as my_col_name" to the approriate columns in the query is pretty trivial, so why don't these user applications correctly do that? Is it genuinely infeasible for some reason? -- Andrew Piskorski <[EMAIL PROTECTED]> http://www.piskorski.com/
