Re: Passing Parameters

From: Tim X <timx_at_nospam.dev.null>
Date: Wed, 27 Aug 2008 18:54:36 +1000
Message-ID: <877ia23kj7.fsf@lion.rapttech.com.au>


sybrandb_at_hccnet.nl writes:

> On Tue, 26 Aug 2008 13:14:29 -0700 (PDT), Mtek <mtek_at_mtekusa.com>
> wrote:
>
>>It just seems totally f**ked up. I mean, I should be able to create
>>a type in every schema and have the database be smart and use the
>>local type when needed, and when arrays are passed, they use the type
>>in the receiving schema.......seems pretty straight forward to
>>me.......
>
> What you 'think' you should be able to do is not important. What is
> implemented and *described* in the documentation, you refuse to read
> ,is.
> Now when you have two identically named types, one 'local' as you call
> it, and one 'global' you are definitely going to create havoc.
> Due to precedence rules the type name is likely resolved only once
> within he procedure.
>
> So it is not f**ked up. You just scr**wed up the whole thing by taking
> the shortcut to the keyboard, without reading and understanding.

Even worse, consider the maintenance nightmare you would be creating! Doing what the OP suggests now means that every developer has to know about *every* definition in every schema and when a change is made, ensure all of them are updated. If any are missed, what sort of error message would be given (assuming it can detect such errors).

Whats the bet that he isn't using %TYPE, %ROWTYPE either?

Tim

-- 
tcross (at) rapttech dot com dot au
Received on Wed Aug 27 2008 - 03:54:36 CDT

Original text of this message