#755: 'make install-dev' assumes "modern" File::Path
----------------------+-----------------------------------------------------
 Reporter:  doughera  |       Owner:       
     Type:  bug       |      Status:  new  
 Priority:  normal    |   Milestone:       
Component:  install   |     Version:  1.2.0
 Severity:  medium    |    Keywords:       
     Lang:            |       Patch:       
 Platform:            |  
----------------------+-----------------------------------------------------

Comment(by doughera):

 Replying to [comment:1 jkeenan]:

 > Now, since I subsequently switched several instances of the
 'traditional' interface to the (no-longer-quite-so) 'modern' interface, if
 you only have 1.06 you should be getting failures in many places other
 than Parrot::Install (''e.g.,'' ''t/configure/*.t'' or ''t/steps/*.t'').

 Actually, those should work ok, though perhaps not for the reason one
 might initially expect.  They use calls of the form {{{ mkpath($path,
 {mode => 0755}).  }}}  The traditional interface expects the second
 parameter to be a boolean (specifying verbosity).  It sees {mode => 0755}
 and evaluates it as true, prints "$path", and then creates the directory
 with the default permissions of 0777 as modified by the user's umask.  The
 test suite ignores the printed message, and the test passes.

 > So if we use the version that comes with what we have designated as the
 minimum version of Perl 5 needed to build Parrot, we cause breakage for
 people who use the latest version from CPAN.  How should we proceed?

 I suspect the problem is probably the t/tools/*.t tests are mis-using the
 traditional interface.
 Enclosing the first argument in square brackets {{{ [  ] }}} should work
 for all versions of File::Path.

-- 
Ticket URL: <https://trac.parrot.org/parrot/ticket/755#comment:3>
Parrot <https://trac.parrot.org/parrot/>
Parrot Development
_______________________________________________
parrot-tickets mailing list
[email protected]
http://lists.parrot.org/mailman/listinfo/parrot-tickets

Reply via email to