Re: Contrib Versions - Mailing list pgsql-hackers

From Tom Lane
Subject Re: Contrib Versions
Date
Msg-id 11233.1305238164@sss.pgh.pa.us
Whole thread Raw
In response to Contrib Versions  ("David E. Wheeler" <david@kineticode.com>)
Responses Re: Contrib Versions  ("David E. Wheeler" <david@kineticode.com>)
List pgsql-hackers
"David E. Wheeler" <david@kineticode.com> writes:
> Hackers,
> I don't suppose I could convince you to use dotted-decimal version numbers for the contrib extension versions, rather
thannumerics, could I? At this point, I think that would just mean changing them from 1.0 to 1.0.0.
 

> Why? Well, PGXN uses semantic versions, which have this format, so I'm biased that way when thinking about dependency
resolution(which is coming in the the PGXN client). But the other reason is because I think it makes sense for all the
versionsin a project to be consistent, and the core versions have (mostly) always used this format.
 

I had somewhat intentionally not numbered them in the same format as the
main release numbers, because if we did that, people would expect them
to match the main release numbers.

I'm also still unwilling to make a core-code commitment to specific
requirements on extension version number format --- we've been around on
that multiple times already, and I don't think the arguments have
changed.

Having said that, I don't really care that much, except that it seems
a bit late in the release cycle to be changing this.  People have
presumably already got installations that they hope to not have to
scratch and reload for 9.1 final.
        regards, tom lane


pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: plpgsql doesn't supply typmod for the Params it generates
Next
From: Alexey Klyukin
Date:
Subject: 'tuple concurrently updated' error for alter role ... set