Drew Taylor wrote:
On Thu, 30 Sep 2004 22:20:25 -0400, Sean Davis <[EMAIL PROTECTED]> wrote:
I had thought about Class::DBI, but the overhead
bothers me a bit
As they say "Premature optimization is root of all evil" :-)
If your code accesses just couple of rows of data, it's safe to guess that overhead of everything else in your code will be higher, than decently designed and maintained module as CDBI.
If (1) you really use thousands of data manipulation calls, and (2) your code is correct, (3) user requirements are stable, and (4) profiling proves you CDBI calls are the culprit, and (5) you cannot solve it by spending $100 on adding hardware, only then IMHO is time to think about spending programmer's time (and organization's development dollars) on custom optimized code. And then, you can use plain DBI and handcrafted optimized SQL to solve it.
So, voice of users of CDBI who know more about the issue than we do:
As for the overhead, it hasn't bothered me (or nearly anyone of the
CDBI list) yet. The maintainer's philosophy is that he's more worried
about getting the functionality right, and then he'll look at
optimizing the queries.
IMHO, IANAL, YMMV, I've never seen your code.
It's better to learn from mistakes of others - I also was lured by premature optimization (because it is more fun), and have to maintain custom data access code in cases where CDBI will do with little or no impact on performance.
-- Peter Masiar
--------------------------------------------------------------------- Web Archive: http://www.mail-archive.com/[EMAIL PROTECTED]/ http://marc.theaimsgroup.com/?l=cgiapp&r=1&w=2 To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
