Re: extension_control_path - Mailing list pgsql-hackers

From Magnus Hagander
Subject Re: extension_control_path
Date
Msg-id CABUevExgnie2O60K0ibt2A0389WjGTZhQh6-7pO-uauU5UC3dw@mail.gmail.com
Whole thread Raw
In response to Re: extension_control_path  (Dimitri Fontaine <dimitri@2ndQuadrant.fr>)
List pgsql-hackers
On Sat, Jan 25, 2014 at 6:07 PM, Dimitri Fontaine <dimitri@2ndquadrant.fr> wrote:
Magnus Hagander <magnus@hagander.net> writes:
> Using colon as the path separator is going to break on windows. The patch
> notices this and uses semicolon on Windows instead. Do we really want to go
> down that path - that means that everybody who writes any sorts of
> installation instructions including this will have to make them separate
> for different platforms. Shouldn't we just use semicolon on all platforms,
> for consistency?

Well, I've been considering that what I found already in the backend to
solve the same problem was a valid model to build against.

Pick any reasonnable choice you want to, fix dynamic_library_path along
the new lines or maybe ask me to, and then let's apply the same design
to the new GUC doing about exactly the same thing?

Ha, I didn't realize dynamic_library_paty had the same problem. In fact, I have to admit I didn't realize I could put more than one path in there - I don't think I've ever used that :D

So based on the previous behaviour there, I withdraw my comment - being consistent with the existing behaviour of that parameter makes perfect sense. 

--
 Magnus Hagander
 Me: http://www.hagander.net/
 Work: http://www.redpill-linpro.com/

pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: [COMMITTERS] pgsql: libpq: Support TLS versions beyond TLSv1.
Next
From: Noah Misch
Date:
Subject: Re: [COMMITTERS] pgsql: libpq: Support TLS versions beyond TLSv1.