Ben Reser wrote on Wed, Jul 03, 2013 at 07:28:56 -0700:
> On Wed, Jul 3, 2013 at 7:20 AM, Daniel Shahaf <danielsh_at_elego.de> wrote:
> > Ben Reser wrote on Wed, Jul 03, 2013 at 07:10:46 -0700:
> >> Not sure how to really handle the Windows build. I guess we can just
> >> add "60" if we test with it because we have no auto-detection, so
> >> you're making a deliberate choice to select a 6.0 version. Maybe add
> >> a warning to the stdout of gen-make.py when you choose to do so?
> > I don't think we need to add a warning that a feature is used that
> > requires a source code patch to be enabled.
> I think you misunderstood what I meant.
> The only argument I see against leaving "60" out of the allowed list
> in the Windows build is that we haven't tested it. Once that changes
> we should allow "60", when we do so we should also add a warning about
> the licensing issues. I say this because Windows has no
> auto-detection of dependencies and as such the version of BDB is
> always deliberately chosen by the user.
We could do the same in configure.ac: permit BDB 6 to go through if it
was explicitly specified in the argument of --with-berkeley-db.
Received on 2013-07-03 16:38:51 CEST
This is an archived mail posted to the Subversion Dev