[Bioperl-l] "progress": useful changes vs. "shiny new thingie"

[Chris Fields]

On Nov 15, 2006, at 6:27 PM, Sendu Bala wrote:

>> I think the old Makefile.PL is still in the 1.5.2 branch.  So  
>> maybe we
>> should stick with that for now (i.e. not change over to  
>> Module::Build on the
>> 1.5.2 branch).  The install script fix that Brian committed to  
>> Makefile.PL
>> in CVS HEAD prior to removal could be merged to branch-1.5.2.
>
> Its certainly possible to stay with Makefile.PL on the 1.5.2 branch. I
> /think/ that if I took a META.yml generated from HEAD's Build.PL and
> added it to the branch that should solve the CPAN issue at least.
>
> I would just want clarification that the consensus really is to stay
> with Makefile.PL for 1.5.2. The primary argument seems to be to not  
> have
> anything too new and untested in the branch, but Makefile.PL itself  
> has
> lots of new additions. My Makefile.PL improvements and the change to
> Build.PL have all been in the name of making 1.5.2 install well. The
> move to Build.PL was just the most appropriate way to fix some bugs  
> and
> make needed changes.

Right, but those changes were made w/o much discussion; what  
discussion there was seemed to end with the agreement this would wait  
until after rel. 1.5.2.  I think that's the main point.  That and the  
timing seems to be a little off for implementing this (after three  
RC's and just prior to a final release).

Everyone seems to think that switching to Module::Build would be a  
good idea, but the move was made rather arbitrarily and (borrowing  
Brian's well-chosen word) unilaterally.  And, although the  
Makefile.PL was a bit clunky, it works for now.

>> We could leave the Module::Build stuff in CVS for the next release  
>> to work
>> out the bugs.  I'm not sure whether we can have both Build.PL and
>> Makefile.PL in CVS HEAD, but I don't see why not.  Wouldn't 'perl  
>> Build.PL'
>> just overwrite the old Makefile.PL anyway?
>
> It would, but why would we want both in CVS? I'd strongly advise the
> cessation of development on the old Makefile.PL, and if it isn't going
> to be added to anymore, nor used by anyone, why keep it in CVS? It's
> just going to cause confusion and problems with people coming across
> issues in the old Makefile.PL. Wasted development time if someone  
> tries
> to fix something in Makefile.PL that was already solved in Build.PL.
> Bizarre behaviour if you were doing something with old Makefile.PL but
> then managed to overwrite it with Build.PL's Makefile.PL and suddenly
> you're working with a whole different script.
>
> In short, not the most optimal of situations.

Stefan has hit upon the problem very succinctly; there is no  
Makefile.PL in CVS, so anyone checking out bioperl-live will  
immediately run into problems.  Hence, there has to be a Makefile.PL  
in CVS HEAD, be it the old one or a new one, at least for the time  
being.  And it seems many still want the old Makefile.PL for now  
(which is why I suggested it).

If you left the old Makefile.PL in you could always add a deprecation  
message indicating the switch.  It will at least give everybody some  
time to get used to using Module::Build and the new system (and  
hopefully dissuade them from working on it).

Chris

Christopher Fields
Postdoctoral Researcher
Lab of Dr. Robert Switzer
Dept of Biochemistry
University of Illinois Urbana-Champaign