NYCPHP Meetup

[Timothy Boyden]

>From my experience the answer is "it depends".
 
If you use the PHP ODBC functions, you'll pretty much be able to
communicate with any database that supports ODBC access without having
to change your code. However you will be restricted to the basic ODBC
functions and database specific optimizations and features will be
unavailable to you.
 
However another approach depending on how much time your willing to put
into it would be to setup your project in a modularized way such that
with a simple configuration change (either via a database or a flat
file) you can direct your application to call pre-defined database
functions based on the configuration setting. So for a MySQL database
you set the configuration to "MySQL" and when your app calls for a
database function it is directed to the specific PHP MySQL database
functions. Obviously this would require writing a lot more code as you'd
have to do this for each type of database you want to support, but if
done properly once you do it you can re-use it for any project you
require it for.
 
I haven't researched it myself, but I'm sure that there are probably
frameworks out there that take care of that functionality for you, but
for me most of them have a lot of features I don't need or have a steep
learning curve for the amount of time I'm willing to put into them.
YMMV.
 
I know my response is vague at best, but then again this is a fairly
common subject that is best answered by you doing a little research via
Google and testing out some of the available solutions to see what works
best for you.
 
-Tim

________________________________

From: talk-bounces at lists.nyphp.org [mailto:talk-bounces at lists.nyphp.org]
On Behalf Of Brian Dailey
Sent: Wednesday, February 14, 2007 10:19 AM
To: NYPHP Talk
Subject: [nycphp-talk] Writing Database Agnostic Code


A lot of the books and other material that I've been reading talk about
database abstraction - developing your application so that you can use
different database backends without significant (or maybe any) changes
to your code. Pear::DB seems to really do a good job with this.
However...

In the past, I've always used MySQL or Oracle, but mostly MySQL. Each
SQL distribution has its own peculiar ways of doing some things, and I
know that using some of the MySQL functions makes the code far easier to
write and also optimizes the speed of the query. 

So my question boils down to this: how does one balance writing code
that works regardless of the backend and still keep things optimized for
speed and clarity? Are there any generalized tips that any of you can
share from your experiences? 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.nyphp.org/pipermail/talk/attachments/20070214/526d324b/attachment.html>