Re: 8.3beta bug or feature? - Mailing list pgsql-general

From Jeff Davis
Subject Re: 8.3beta bug or feature?
Date
Msg-id 1199995334.10057.428.camel@dogma.ljc.laika.com
Whole thread Raw
In response to Re: 8.3beta bug or feature?  (marcelo Cortez <jmdc_marcelo@yahoo.com.ar>)
List pgsql-general
On Thu, 2008-01-10 at 16:19 -0300, marcelo Cortez wrote:
> > The cast already exist; I think you could change its
> > context (from
> > "explicit" to "assignment" IIRC).  This is, of
> > course, not recommended.
>
>
>    I've created cast with  assignment from in4 to text
>
>  but
> select 23  LIKE '2%'  fail.
> ERROR:  operator does not exist: integer ~~ unknown
> LINE 1: select 23  LIKE '2%'
>                    ^
> HINT:  No operator matches the given name and argument
> type(s). You might need to add explicit type casts.

First of all, I think Alvaro meant "change to implicit", I don't think
"assignment" will help.

Second, assuming that works for that specific case (which it may or may
not, I haven't tried it), it's still not a good solution for you.
Implicit type coercion causes all kinds of subtle problems that are hard
to see, which is why the change was made in 8.3.

And if I understand correctly, just changing a few casts to "implicit"
will not result in behavior the same as 8.2, so it's likely to break
other parts of the application.

I don't think there's any way around fixing the application. You can
still use 8.2 with tsearch2 (which still work great) until the mapper is
fixed.

What is the name of the third-party product?

Regards,
    Jeff Davis


pgsql-general by date:

Previous
From: "Isak Hansen"
Date:
Subject: Re: 8.2.4 serious slowdown
Next
From: Clodoaldo
Date:
Subject: Re: ATTN: Clodaldo was Performance problem. Could it be related to 8.3-beta4?