[Phoenix-pm] PERL DBI

Scott Walters scott at illogics.org
Mon Mar 13 13:46:18 PST 2006


On  0, "David A. Sinck" <phx-pm-list at grueslayer.com> wrote:
> 
> HAHAHAHAHA.  Guess again.  There's an instance, at least with MySQL

Okay, I don't use MySQL heavy (for reasons of it better garbage), so 
I over generalized.  I conceede your point that rare bugs will make
this generalization not always true.

> \_ and you don't suffer
> \_ from SQL command length limitations, 
> 
> I bet I could could craft something that hits a buffer still using
> placeholders if I wanted.  :-)

I didn't say he didn't suffer from *buffer* length considerations -- that
would be foolish indeed.  I only said that he doesn't suffer from this
particular one.  I've seen databases cap the command buffer at 64k.
Generally speaking (true in most cases, perhaps false in some) the 
length-counted buffers for placeholder data will be significantly larger
than the command buffer.

> 
> \_ and the database doesn't have to parse potentially
> \_ megs of data just to parse the SQL command, 
> 
> Depends on the driver.  MySQL dbd, last I checked, actually sub'd the
> values in before it got passed to the engine.  YMMV.

That must have been a long time ago indeed.  But you caught me --
I made another false generalization.  I should have said "Generally, "
in front of it. 

By the way, people who argue small points while ignoring the point of
what was said annoy me.  Did you really think someone on the list
would be damaged thinking that placeholders always send data out of band,
even when using Informix?

I wrote a hasty reply out of desire to be helpful to someone who was
having trouble with something, not to debate minutia, and have been
punished for it.

-scott

> 
> David
> _______________________________________________
> Phoenix-pm mailing list
> Phoenix-pm at pm.org
> http://mail.pm.org/mailman/listinfo/phoenix-pm


More information about the Phoenix-pm mailing list