Jonathan, while you are well-meaning in your comments, you are
mis-reading what I have said multiple times and are therefore making
a straw man argument against it.
Regarding this point:
>5. All details used to construct a connection handle should be
completely decomposed rather than shoved into an ungainly "data
source". Examples of what should be distinct (not all being
applicable at once) are: <snip>
That's the key part of my proposal right there. I don't care about
exactly what break-down of pieces we collectively come up with, as
long as they are unambiguous, reasonably atomic, account for all the
possible needs, and have reasonable names. That 11-point list of
mine was only meant to be an example of what breaking up a "data
source" string can give us; it was not meant to be exhaustive and
there is no point in arguing at length against what it missed and how
horrible it is. No doubt, an IDS and/or components thereof should
also be in the list of connection parameters. I'm not arguing
against what one needs for Informix, either explicitly or implicitly.
All this said, thank you for all the information about Informix how
you set up your systems, and how many levels of hierarchy there can
be with a database; eg, a server has 1+ databases, each of which has
1+ catalogs (and presumably each of those has 1+ schemas and each of
those 1+ tables and other schema objects). I will probably find it
helpful when I plan various DBI-using tasks.
-- Darren Duncan
