Hi,
Am Mittwoch, den 01.12.2004, 16:46 +1100 schrieb Russell Smith:
> Regarding the "Natural Order" of columns.
>
> Why if we can delete a column from the middle of a table now, change the type
> of that column to something totally different, eg text -> int.
> Can we not move the order of the rows around, and when the new row is written
> to disk in the new arrangement. Or more accurately,
> why is it not possible to add a new column, not at the end of the list. It's
> probably more complicated that I think, as that's usually the case.
How do you select "middle of a table" ?
All I know is how to refer to columns by name. And for that it is
unimportant in which order they appear in SELECT *
In fact its not recommendet to use SELECT * in production code.
And again, SELECTS on one table only are very rare in most
projects I've seen - so why pull any extra column you arent
going to use in an app over the wire?
There are edge cases when you want to write a generic database
tool where you just display data as it is, but then you could
easily maintain your own "order of columns" in a table.
Usually a table even has some columns with keys, would you
show that to a user? And if so, why?
Some even dont use a single table here because they dont feel
comfortable let the average user fiddle with the schema.
So whats the point to call for that "cosmetic" feature
again and again?
Just my 2ct
Tino
---------------------------(end of broadcast)---------------------------
TIP 6: Have you searched our list archives?
http://archives.postgresql.org
