develooper Front page | perl.perl5.porters | Postings from June 2013

Re: is dead, long live

Thread Previous | Thread Next
Peter Scott
June 2, 2013 01:00
Re: is dead, long live
Message ID:
On Sat, 01 Jun 2013 10:02:18 -0400, Ricardo Signes wrote:
> * Peter Scott <> [2013-06-01T09:20:24]
>> I would like to add a belated argument for the pedagogical value of
> I am a *big* fan of pedagogical arguments!
> That said, this sounds to me like an argument not to delete from
> the CPAN, rather than to not remove it from core.  Is the issue that
> students must (fsvo "must") only be taught with things installed in the
> core distribution in many contexts?

Not a must.  This isn't cut-and-dried.  Of course at some point the 
students must learn how to install from CPAN.  It is not a short 
conversation.  Not (for the most part) because we don't have succinct 
tools but because of the decision points: local::lib vs root and the RPM 
option, cpanm vs modules that need manual intervention.  And at some 
point that arrives all too soon, they will grab some innocent module that 
downloads 50 others, one of which requires some sleuthing.

So I prefer to introduce as much value as possible and get the student to 
a certain degree of comfort with Perl before subjecting them to that so 
that they trust that the effort is going to be worth it.  Can I move from one side of that time-to-learn-CPAN line to the other?  Of 
course.  I just happen to feel that its usefulness to the beginner trying 
to get a handle on this stuff justifies what seems to be a small cost of 
keeping it in the core, because it can be used and understood with the 
relatively little knowledge and expertise of students who haven't reached 
the line yet.  Consider also not a little legacy documentation of various 
kinds that expects to be in core, all of which could be considered 
an argument for backwards compatibility.

Peter Scott

Thread Previous | Thread Next Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at | Group listing | About