On 1/11/08, Samuel R. Neff <[EMAIL PROTECTED]> wrote: > > The same could be said for pretty much any other database.. they're all > similar 'cause they all follow (to some extent) the same standard. If > aliases were defined for PostGres then why not for MySQL, Oracle, MSSQL, > Firebird, VistaDB, SAP/DB, DB/2, and on and on.
Good point. But since I am only interested in either SQLite or PostGres in the open source world (as of now, the only world that interests me for the most part), I asked specifically about SQLite --> PostGres compatibility. > > I don't agree that defining aliases into the standard distribution would be > a good thing, but it would be nice to have a way to define aliases that were > then handled and substituted by SQLite, so people can have a list of aliases > that each can load and then run the SQL dialect of their choice. There is a > bigger problem though in areas that an alias can not handle, like keywords > that have to be placed in different parts of SQL (LIMIT vs TOP) or operators > that have different meaning (+ vs ||). > well, yes, I don't care how it is done... if providing a mechanism for creating user-defined aliases is cheaper than creating the actual aliases compatible with a specific other database, then so be it. I speak all this from the vantage point of someone who knows next to nothing about the innards of SQLite (am not a C person) but fully understands and respects the philosophy of the makers of SQLite. I am just trying to see how much that philosophy can allow while making SQLite into a great app development platform that can seamlessly be upsized to a major database such as PostGres. > > -----Original Message----- > From: P Kishor [mailto:[EMAIL PROTECTED] > Sent: Friday, January 11, 2008 2:19 PM > To: sqlite-users@sqlite.org > Subject: [sqlite] SQLite --> PostGres > > I have been writing an app using SQLite (great to develop with, and to > serve with to an extent), always with the intent of one day upsizing > it to PostGres (pain in the tush to develop with, but great to serve > with to any extent). Much to my delight, I am finding that y'all (the > SQLite developers) have made many things (for example, datatypes) > similar to PostGres (yes, I know most all about how SQLite datatypes). > My question -- why not take it all the way, and make SQLite almost a > mini-PostGres... wait, before you chide me -- I don't mean in the "add > more features" way, but in the "make syntax and datatypes as similar > as possible" way. > > For example, why have the "INTEGER PRIMARY KEY" when it could just as > easily be called "SERIAL"? > > One way might be to allow for aliases -- so, SERIAL would mean the > same as INTEGER PRIMARY KEY, or CHAR(3) would mean the same as TEXT > with a CHECK CONSTRAINT, and so on. > > Wouldn't that increase the already wildly popular appeal of SQLite > even more so without subtracting or weighing anything down? > > By the way, I didn't find a BLOB kind in PostGres -- is that the same > as BYTEA? If yes, that would be another candidate for such an alias. > > > -- > Puneet Kishor > ----------------------------------------------------------------------------- To unsubscribe, send email to [EMAIL PROTECTED] -----------------------------------------------------------------------------
