Re: Missing file versions for a bunch of dll/exe files in Windows builds

On Thu, Apr 24, 2014 at 10:18 AM, Dave Page <dpage@pgadmin.org> wrote:

> On Thu, Apr 24, 2014 at 9:13 AM, Magnus Hagander <magnus@hagander.net>
> wrote:
> > On Thu, Apr 24, 2014 at 9:50 AM, Dave Page <dpage@pgadmin.org> wrote:
> >>
> >> On Thu, Apr 24, 2014 at 8:27 AM, Magnus Hagander <magnus@hagander.net>
> >> wrote:
> >> > On Thu, Apr 24, 2014 at 7:54 AM, Michael Paquier
> >> > <michael.paquier@gmail.com>
> >> > wrote:
> >> >>
> >> >> Hi all,
> >> >>
> >> >> A colleague of mine pointed out that the file version is missing in a
> >> >> couple of dll and exe files when building on windows using the
> >> >> community scripts in src/tools/msvc. After having a closer look, I
> >> >> noticed that a *lot* of files are missing the shot:
> >> >> - all the exe/dll in contrib/
> >> >> - dll of PL languages (perl, python, tcl, pgsql)
> >> >> - libpqwalreceiver, snowball
> >> >> - ecpg stuff
> >> >> - regression and isolation test stuff
> >> >> - conversion_procs thingies
> >> >> Having a version number associated to a build is important for all
> >> >> companies creating builds of Postgres on Windows, so it would be good
> >> >> to have the patch attached applied and back-patched.
> >> >>
> >> >
> >> > At least some fo that is intentional - things that are considered
> >> > "internal"
> >> > were not given a version resource intentionally. E.g. the
> >> > conversion_procs
> >> > is very intentional.
> >>
> >> Why? Version resources are kinda handy for installers to reliably
> >> figure out whether or not something should be upgraded.
> >
> >
> > I think the general argument was "maintenance overhead". This was back
> > before the 8.0 release, so my memory is somewhat sketchy :)
> >
> > Which also means of course that we can reconsider that decision. But if
> we
> > do, we need to support it equally in mingw and msvc.
>
> Well it's pretty darn close to zero maintenance isn't it? We don't
> even have to maintain rc files with the proposed patch (if my
> extremely weak perl-fu is reading the code correctly).
>
>
Probably, yes. It may just have been the caution of the first release..

That said, as also noted, the patch doesn't actually fix the problem :P For
one it establishes two different ways of generating the resources. And it
also doesn't support mingw. So it's likely going to be a little bit more
involved, but probably not hugely much. And if we do it right, ongoing
maintenance should hopefully be very low.

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