Re: User Developed ODBC

From: Richard Armstrong <rarmstro_at_ix.netcom.com>
Date: 12 Oct 1994 03:31:45 GMT
Message-ID: <37flb1$o7f_at_ixnews1.ix.netcom.com>


In <37dv8g$duu_at_carbon.denver.colorado.edu> ghharrac_at_ouray.Denver.Colorado.EDU (George) writes:

>
>HI,
>
>Has anyone developed their own ODBC like interface using PRO*C method 4?
>or something similar wuth OCI. If you have, have you been able to extend
>your master piece to support multiple cursors. I'am having problems on
>the support of multiple cursors ( ODBC cannot support this) and I am looking
>for someone to disguess common problems with.
>
>ghharrac_at_ouray.denver.colorado.edu
>
>
>
>
>

I have written my own ODBC. I had to use OCI to do it since Pro*C is not able to open dynamic cursors. (You must have a blasted name for each cursor at compile time in Pro*C) OCI is much easier to use when writing an ODBC (actually, it is the only tool that will work). The only thing that made me wonder is: why are there so many ODBC calls? and why does it take so many ODBC calls to do so little work? Most of my ODBC functions don't do anything... primarily because there are not that many steps to binding to variables, describing, and executing SQL statements. I really don't know why Microsoft made ODBC so complicated!

The other nice thing about using OCI is that OCI calls more closely resemble the functionality ("touch-and-feel") of ODBC calls.

-Richard Received on Wed Oct 12 1994 - 04:31:45 CET

Original text of this message