[Bioperl-l] hash randomization failures for bioperl modules (perl 5.18.0)

[Fields, Christopher J]

A quick update: many of the bugs have been fixed with the exception of two: one is from Bio::Map::Physical and the other from Bio::Assembly::Tool::ContigSpectrum.  I have TODO'd the relevant tests to indicate they have problems, but I'll file a bug on these to indicate more work needs to be done.

chris

PS: latest travis-ci, all tests passing now for perl 5.18:

    https://travis-ci.org/bioperl/bioperl-live/jobs/10074646

On Aug 9, 2013, at 3:24 PM, "Fields, Christopher J" <cjfields at illinois.edu> wrote:

> Just a note that several modules are failing for perl 5.18.0; this is likely due to the new hash randomization added:
> 
>    http://search.cpan.org/~rjbs/perl/pod/perldelta.pod#Hash_randomization
> 
> Travis-CI test results are here:
> 
>    https://travis-ci.org/bioperl/bioperl-live/jobs/10041002
> 
> I can confirm these as well running them locally.
> 
> It's possible these are relegated to the tests themselves, which I'll check into.  I have also committed a fix for the Bio::Map tests.  In the meantime, just a note for the future: these are true bugs in the sense that there is some reliance on the order of keys/values returned from a hash remaining the same (e.g. reliance on the hash seed not changing).  This is not the case and has long been noted in perl core documentation, so the perl devs have introduced hash randomization to pinpoint these problem areas.  
> 
> chris
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/bioperl-l