Re: Adding optionally commit number in PG_VERSION_STR - Mailing list pgsql-hackers

From Tom Lane
Subject Re: Adding optionally commit number in PG_VERSION_STR
Date
Msg-id 3038.1374031659@sss.pgh.pa.us
Whole thread Raw
In response to Adding optionally commit number in PG_VERSION_STR  (Michael Paquier <michael.paquier@gmail.com>)
Responses Re: Adding optionally commit number in PG_VERSION_STR
List pgsql-hackers
Michael Paquier <michael.paquier@gmail.com> writes:
> It happens that I work occasionally on multiple builds based on
> different stable branches at the same time to check fixes that need to
> be backpatched, and I tend to easily lose track on which version the
> build I created is based on (Duh!). There is of course the version
> number up to the 3rd digit available (for example 9.2.4, 9.3beta2,
> etc.), but as a developer I think that it would be helpful to include
> the commit ID in PG_VERSION_STR to get a better reference on exactly
> what the development build is based on. This could be controlled by an
> additional flag in ./configure.in called something like
> --enable-version-commit, of course disabled by default. If enabled,
> PG_VERSION_STR would be generated with the new information. configure
> would also return an error when this flag is enabled if git is either
> not found, or if the repository where configure is not a native git
> repository.

Personally, I'd find that pretty useless, because the build I'm running
has typically got uncommitted changes in it.  Also, there isn't any good
way to set PG_VERSION_STR except at configure time, which means the
common practice of doing (at most) "make clean" before rebuilding would
not result in an updated string even if you had committed.
        regards, tom lane



pgsql-hackers by date:

Previous
From: Andreas Karlsson
Date:
Subject: Re: Listen/notify across clusters
Next
From: Ants Aasma
Date:
Subject: Re: Improvement of checkpoint IO scheduler for stable transaction responses