Eric Wilhelm wrote:
# from m.nooning
# on Thursday 25 January 2007 11:41 am:
I replaced the one in C:\perl\site\lib\App\Packer\PAR.pm with the new
one and pp would not work at all, so that rules out interference from
that paticular PAR.pm. I don't see that any of the others could
interfere.
That's not the PAR.pm. That's App/Packer/PAR.pm. You want to look at
these two:
C:\perl\site\lib
C:\perl\cpan\build\PAR-0.970\lib
Your OS is qdos? That worked on the Intel 8086! I assume you have
upgraded your hardware since then but, wow!
That's a joke. Windows XP is still qdos, no matter how shiny.
--Eric
I sync'd the cpan par file with the new site\lib one. The unbundling
/execution took times of 13, 10 and 13 seconds. Actually, no OS can be
counted on for consistency in these matters. A system need could occur
any time and take up resources during the unbundling.
As long as you are using XP, let me switch subjects to the command line
-c switch problem.
Does what is pasted below happen to you? The test for it works in the
environment of the "nmake test" prior to installing a new PAR-Packer.
Still, the resultant "a.exe" should not fail when run from the command
line. It hints of an $ENV var set during the "official" tests, but not
set in the tests below. Printouts of %ENV placed into the pp.pm file
did not reveal anything suspicious.
---------------------------paste
C:\aaa>type hello.pl
print "hello\n";
C:\aaa>pp -c hello.pl
Set up gcc environment - 3.4.2 (mingw-special)
C:\aaa>a
mkdir
C:\DOCUME~1\malcolm\LOCALS~1\Temp\par-malcolm\cache-5dc7a7f33d3db9d331e9756a16e18cdce6436902/inc/lib/C:/
: Invalid argument at C:/perl/lib/Archive/Zip.pm line 659
C:\aaa>a
hello
------------------end paste
The -c switch is a pretty low priority bug, and it is not trivial, as
prior emails indicate. In any case for the few two weeks I will be
busier than usual so I am going to put it on a back burner for now. I
will still be checking things from time to time though.
Thanks
