[Bioperl-l] No more Makefile.PL (MakeMaker -> Module::Build)
cjfields at uiuc.edu
Thu Nov 16 10:29:31 EST 2006
> > Is there a reference for the format anywhere?
> Erm not really - that I know of. There is this thread which I use for
Looks like kobes is using PPM::Make (he should; it's his distribution).
We'll need to make sure that Module::Build uses that as well, since it seems
there are some significant changes.
> There must be an automated way to generate the PPM4
> compatible PPDs - I think there is a way to generate one from
> the CPAN distribution. I think this might be the way
> ActiveState does it?
I think ActiveState has it's own automated build process. For PPM3 they
used the PPD section from the regular makefile (from Makefile.PL and
ExtUtils::MakeMaker). Notably they still indicate using 'nmake ppd' when
generating Makefile.PL, so maybe they use an updated or modified
ExtUtils::MakeMaker? Or they add tags for the Makefile.PL:
# See lib/ExtUtils/MakeMaker.pm for details of how to influence
# the contents of the Makefile that is written.
'NAME' => 'Term::Control',
'VERSION_FROM' => 'Control.pm', # finds $VERSION
($] ge '5.005') ? (
'AUTHOR' => 'Johnny Doel (johnny at doel.org)',
'ABSTRACT' => 'Control the IO for terminals',
# extra tags here?
) : (),
The reason you may not see a certain distribution built into a PPM is b/c a
particular dependency is not available for one reason or another, tests fail
during the make process, etc. They will try to build specific ones upon
request. Many of the ones in Kobes are generated by users who want
up-to-date distros or ones not built automatically by ActiveState, though
Randy Kobes himself sets up most of them. Significantly, he has already
switched over to PPM4; maybe we should contact him about this?
Postdoctoral Researcher - Switzer Lab
Dept. of Biochemistry
University of Illinois Urbana-Champaign
More information about the Bioperl-l