Re: !! Pro*C versus OCI !!

From: Joel D. Elkins <jdelkins_at_iquest.net>
Date: Thu, 5 Jan 1995 02:32:41 GMT
Message-ID: <jdelkins-0401952132420001_at_ts1-ind-18.iquest.net>


In article <3ec5gd$el6_at_ixnews2.ix.netcom.com>, bxl_at_ix.netcom.com (Bill Lo) wrote:

>Question: Is there any good C++ class library that wraps OCI exists
>under Windows/NT/OS2 ?
>
>It seems to me a C++ wrapper will be the best approach.

In my current project, our requirements were for a C++ Oracle class library, usable on Macintosh and Win16 and Win32s. We opted to roll our own, and we're very happy with the results. Our needs are relatively simple--our data model must be relational (can't use object persistence because of corporate data model standards)--and if your's are too, you might investigating rolling your own classes as we have. Our implementation is proprietary, so I can't disclose the source, but I might be able to point you in the right direction if you're interested.

Advantages of rolling your own:
1.) Fast - thin layer over OCI + array binds = speed. 2.) Cheap - simple implementation (cross platform) might take as little as 250

    man-hours.
3.) Optimal - Does exactly what you want and no more. Small footprint.

Advantages of buying 3rd party library:

1.) Easy - probably well-thought-out and intuitive to use.
2.) General - one-stop shopping. May even support other RDBMSs.
3.) Simple - a comprehensive DB class lib will likely simplify the rest of

    your OO model. In our case we had to make a few compromises to such     simplicity.
4.) Comprehensive - our classes basically allow SQL and PL/SQL execution and

    generalized "record-set" transaction processing. A store-bought lib will     give you more.

Good luck.
Joel

-- 
____________________________________________________________________________
Joel D. Elkins N5USU   'The particular sequence of ten words "attack
jdelkins_at_iquest.net    on an English writer that the character
Indianapolis, IN       of this" is not at all unreasonable.'--Claude Shannon
PGPprint = 2C 1F 82 04 36 98 E2 BC  7C B0 5D 9B 39 59 8B E5
____________________________________________________________________________
Received on Thu Jan 05 1995 - 03:32:41 CET

Original text of this message