Re: [HACKERS] RMT: Use Visual Studio 2015 for Compiling and linkingthe Windows version in PG10 - Mailing list pgsql-hackers

From Robert Haas
Subject Re: [HACKERS] RMT: Use Visual Studio 2015 for Compiling and linkingthe Windows version in PG10
Date
Msg-id CA+TgmoazEi91a35jXoeLYNkfYbzcKgLRqG1ii3LLs_J_D9qp1w@mail.gmail.com
Whole thread Raw
In response to [HACKERS] RMT: Use Visual Studio 2015 for Compiling and linking the Windows version in PG10  ("Hans Buschmann" <buschmann@nidsa.net>)
List pgsql-hackers
On Mon, Apr 10, 2017 at 11:21 AM, Hans Buschmann <buschmann@nidsa.net> wrote:
> On 03/28/2017 06:46 AM, I wrote in pgsql-general:
>> I use Postgres on Windows 64 bit (8.1,10,WS2012R2,WS2016) for quite a
>> while.
>> I always install and upgrade from the ZIP binary distribution from
>> enterpriseDB which works like a charm.
>>
>> In a recent fresh install I noticed that PostgreSQL 9.6.2 (older
>> versions not tested anymore) requires Visual C++ 2013 Runtime library to
>> be installed.
>>
>> It seems all binaries are linked against msvcr120.dll (verified with
>> ProcExplorer of Sysinternals).
>>
>> I propose to switch to Visual Studio 2015 for the next distribution:
>> - no more install of > 5 years old runtime library
>> - seems to need no extra runtime library on newer systems (there is no
>> msvcr140.dll after installation of VCR 2015)
>> - when needed, VC Runtime Library 2015 is the same requirement as for
>> recent Apache 2.4.xx versions
>> - VS2015 build is already member of the build farm
>> - I could easyly test the beta on recent Windows 64 OSs
>>
>> BTW, nowhere in the current documentation is mentioned the requirement
>> of the VC++ runtime library installation.
>>
>> Thanks in advance
>
> Now after feature freeze of the new version I wanted to address this topic
> to the Release Management Team.
>
> Opinions?

The one response you got on pgsql-general suggests that this isn't a
proposed change to PostgreSQL but rather a proposed change to
EnterpriseDB's packaging of PostgreSQL.  The RMT has no jurisdiction
over EnterpriseDB's packaging of PostgreSQL, and I have no personal
knowledge of that topic.  You could contact Dave Page about that
issue, perhaps.

If there is a change to PostgreSQL itself proposed, the RMT could
consider granting an extension to feature freeze for that change if
the larger PostgreSQL development community felt that the change was
worthwhile, urgent, and low-risk.  However, the RMT would probably not
want to make such a decision about the matter without hearing from
those community members who know something about Windows, or in
advance of the patch having been written.

-- 
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company



pgsql-hackers by date:

Previous
From: Magnus Hagander
Date:
Subject: [HACKERS] pg_upgrade vs extension upgrades
Next
From: Aleksander Alekseev
Date:
Subject: Re: [HACKERS] GCC 7 warnings