On Fri, Mar 05, 2021 at 05:24:25PM +0000, Ed . wrote:
> > > Whenever you switch lib (or
> > > indeed Perl version), you’d need to re-run a build, because your Perl
> > > is now different.
> > But I didn't change the perl version, only the PDL version. I was in
> > the process of rebuilding/testing/installing Photonic.
>
> Sure, but I did say “or” above

Right. Sorry.

> – and you didn’t just change the PDL version, you changed the lib. Changing 
> lib changes what Perl libraries are available to the “perl” you run on the 
> command line.
I guess I didn't change the lib, I added a lib to the front of the
@INC array, so all the previously installed packages were available
unless those I explicitly reinstalled.

> That means you need to re-run the build from scratch (which I didn’t spell 
> out before, sorry): perl Build.PL; ./Build (or perl Makefile.PL; make).

That's what I was doing. ./Build realclean; perl Build.PL; ./Build;
./Build test;

>
> > I got the errors when Build-testing the Photonic
> > package. Unfortunately, the only error message I got was that the
> > planned number of tests were not run. I tried 'proving' individually
> > the test programs (i.e., prove t/...; prove -v t/...) but they
> > messages were not much more informative. I tried running the tests
> > directly (perl t/...) but it gave me a segmentation fault.
> > So maybe I can rephrase my question in a more general way: how do you
> > debug test programs (programs that use Test::More) to find out
> > why they fail.
>
> Great question! After re-running the complete build (as above), you can run 
> an individual test-file with “prove -b t/file.t” – the “-b” tells it to use 
> the local “blib”, which is where the built files end up. To mimic that and 
> get the truly raw output (not as interpreted by the harness), you’d go “perl 
> -Mblib t/file.t”. I don’t know why you’d get a seg-fault, sorry.

Thanks for these suggestions: I'll use them next time. I thought that
you could not simply feed the test files to perl, but I don't recall
why I got that idea; maybe I tried it at some time in the past
and got and error which I couldn't figure out, surely for not
including the -Mblib switch.

Thanks and regards,
Luis



--

                                                                  o
W. Luis Mochán,                      | tel:(52)(777)329-1734     /<(*)
Instituto de Ciencias Físicas, UNAM  | fax:(52)(777)317-5388     `>/   /\
Av. Universidad s/n CP 62210         |                           (*)/\/  \
Cuernavaca, Morelos, México          | moc...@fis.unam.mx   /\_/\__/
GPG: 791EB9EB, C949 3F81 6D9B 1191 9A16  C2DF 5F0A C52B 791E B9EB


_______________________________________________
pdl-devel mailing list
pdl-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/pdl-devel

Reply via email to