[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Bug#216492: FTBFS (unstable/all) missing build-dep



On Sun, Oct 19, 2003 at 08:15:43PM +0100, Andrew Suffield wrote:
> On Sun, Oct 19, 2003 at 01:20:39PM +0100, Scott James Remnant wrote:
> > On Sun, 2003-10-19 at 09:51, Adam Conrad wrote:
> > 
> > > Package: libtool
> > > Version: 1.5-3
> > > Severity: serious
> > > 
> > > libtool fails to build from source on all the buildds[1] due to a missing
> > > build-dep on texi2html.
> > > 
> > libtool (and libtool1.4) *have* a build-dep on texi2html (and texinfo):
> > 
> > Build-Depends-Indep: debhelper (>= 4.0), texi2html, texinfo
> > Build-Depends: debhelper (>= 4.0), file, g77 | fortran77-compiler, gcj [!hppa !mips !mipsel]
> > 
> > My reading of policy suggests that this is correct:
> > 
> >      `Build-Depends-Indep', `Build-Conflicts-Indep'
> >           The `Build-Depends-Indep' and `Build-Conflicts-Indep' fields must
> >           be satisfied when any of the following targets is invoked:
> >           `build', `build-indep', `binary' and `binary-indep'.
> >  
> > [1]  If you make "build-arch" or "binary-arch", you need Build-Depends.  If
> >      you make "build-indep" or "binary-indep", you need Build-Depends and
> >      Build-Depends-Indep.  If you make "build" or "binary", you need both.
> > 
> > texinfo and texi2html are used in the "build" target.  As far as I can
> > tell this means that the buildd should be ensuring both Build-Depends
> > and Build-Depends-Indep are installed before running it.
> 
> Other people have covered why this breaks. Here's the solution I use:
> 
> Make your build target do nothing.
> 
> That is, make build an empty target that does _not_ depend on
> build-arch and build-indep. Then make sure that binary-{arch,indep}
> will result in the right things getting run anyway.
> 
> This a) preserves the desired effect of the time consuming arch-indep
> stuff not being run on the buildds, and b) actually works. While it's
> not strictly in accord with policy as written, it has the advantage of
> doing what policy expected to happen, and I've never seen a better
> idea.
> 
> Ultimately we should either deprecate the build* targets, or make
> build-{arch,indep} mandatory and deprecate build.

What needs to happen to get this into policy?

Chris

Attachment: signature.asc
Description: Digital signature


Reply to: