dolfin team mailing list archive

[Anders Logg <logg@xxxxxxxxx>]

On Sat, Apr 25, 2009 at 07:33:08AM +0200, Johannes Ring wrote:
> On Sat, April 25, 2009 07:22, Garth N. Wells wrote:
> >
> >
> > Anders Logg wrote:
> >> On Fri, Apr 24, 2009 at 08:53:50PM +0200, Johannes Ring wrote:
> >>> On Fri, April 24, 2009 18:56, DOLFIN wrote:
> >>>> One or more new changesets pushed to the primary dolfin repository.
> >>>> A short summary of the last three changesets is included below.
> >>>>
> >>>> changeset:   6026:b30965648ff4b42614058960be4e3e413e065f32
> >>>> tag:         tip
> >>>> user:        "Garth N. Wells <gnw20@xxxxxxxxx>"
> >>>> date:        Fri Apr 24 17:56:13 2009 +0100
> >>>> files:       SConstruct scons/simula-scons/simula_scons/__init__.py
> >>>> description:
> >>>> BoolOption -> BoolVariable to precent depracation warning.
> >>> This change does not work with the SCons version that is in Hardy:
> >>>
> >>> http://fenics.org:8080/builders/dolfin-hardy-i386/builds/233/steps/dolfin%20build/logs/stdio
> >>>
> >
> > I thought that this might happen.
> >
> >>> Should we require a newer SCons?
> >>>
> >>> Johannes
> >>
> >> That would be ok for me, but there may be other opinions.
> >>
> >
> > OK with me too. There is another deprecate warning that it would be nice
> > to get rid of
> >
> >      scons: warning: The PathOption() function is deprecated; use the
> > PathVariable() function instead.
> >
> > but this may require a simula_scons change.
> 
> OK, which version should be the new required version? 0.98.5, the one in
> Intrepid?
> 
>     EnsureSConsVersion(0, 98, 5)
> 
> ?
> 
> Then we can finally use GetBuildFailures to check for build failures.

That would be *really* good.

It's a bit funny now that the build system does not know if the build
failed: Hey, I'm the build system and I don't know if this worked or
not but look at the lines above for signs of error messages. :-)

-- 
Anders

Attachment: signature.asc
Description: Digital signature