Re: search_path vs extensions - Mailing list pgsql-hackers

From Andrew Dunstan
Subject Re: search_path vs extensions
Date
Msg-id 4A20506C.3070009@dunslane.net
Whole thread Raw
In response to Re: search_path vs extensions  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers

Tom Lane wrote:
> Greg Stark <stark@enterprisedb.com> writes:
>   
>> I'm actually not sure if we should allow extensions to be installed
>> into separate schemas.
>>     
>
> It's starting to seem that best practice is to install "public"
> functions/etc into a common schema and "private" objects into an
> extension-specific schema.  The main problem with that from an extension
> author's point of view is the need to explicitly qualify all references
> to private objects, since they won't be in the search path.  Which is
> tedious, but doable.
>   

The main problem as I see it is that you are abandoning one of the two 
uses of schemas, namely namespace separation.  With this pattern an 
extension author has no guarantee that there won't be a name collision 
with some other extension. Pace Greg, schemas are not just about privacy.

cheers

andrew




pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: search_path vs extensions
Next
From: Hannu Krosing
Date:
Subject: Re: Python 3.0 does not work with PL/Python