Re: OCI vs. Pro*C

In article <2gq919$1d6_at_access3.digex.net> mstein_at_access3.digex.net (Michael P. Stein) writes:
> I find Pro*C easier for static SQL, but for fully dynamic SQL (especially
>Pro*C "Method 4") I find OCI easier.

This is actually the reason why I posted the original question. In my current application we have a library of 17-odd functions or so (like SelectSet, ReplaceSet, CancelQuery, GetLastErrText, etc) which provide a vendor-independent database interface to the rest of our applications. SelectSet and ReplaceSet take a single SQL statement as a null-terminated string, so obviously the supporting code has to do fully dynamic SQL. It seems quite clear from the messages posted here (and from the private e-mail I've received) that OCI is the way to go to re-implement these functions for Oracle.

I even got a phone-call from an Oracle person who asked where I'd gotten the notion that pro*c was depracated. I can't honestly say, but somewhere back in the misty reaches of my mind, I recall that the last time I looked at Oracle (back when the current version was 5.x), that some Oracle marketing person told me, in response to a question about the availability of a 3gl interface, that OCI was available but that Oracle encouraged customers towards pro*c. Maybe I perceived FUD about OCI where there wasn't any.

Thanks for the vigorous dialog on the topic. It's been really helpful.

-- 
Mark Lawrence                                            mark.lawrence_at_drd.com
DRD Corporation                                          +1.918.743.3013
5506 South Lewis Ave.                                    +1.918.745.9037
Tulsa OK 74105