Re: DBI-git on perl-5.8.8

2017-09-21 Thread Tim Bunce
On Mon, Sep 18, 2017 at 07:25:54PM +0200, H.Merijn Brand wrote:
> On Mon, 18 Sep 2017 17:57:39 +0100, Tim Bunce 
> > 
> > > This is fixed by updating Storable, so the minimum requirement for
> > > Storable's version is too low (or in our case, not required at all)
> > > 
> > > perl-5.8.8 with Storable 2.15 failed
> > > perl-5.8.8 with Storable 2.51 passes
> > > perl-5.8.9 with Storable 2.19 passes
> > > 
> > > So I'd suggest to make Storable version 2.19 a minimal requirement:  
> > 
> > Typo here, I think you meant 2.16 not 2.19.
> 
> I meant 2.19, as that is the first version after the failing 2.15 that
> appeared in a stable perl. That implies that it is much harder to test
> against 2.16 than it is to test against 2.19
> 
> Additionally, Storable 2.51 installs flawless on 5.8.8, I I made the
> safer jump from 2.15 to 2.19 instead of to 2.16 which I did not test
> 
> If you agree with 2.19 because of this reasoning, I'll commit that
> change with some comment

Yes, thanks!

Tim.


Re: DBI-git on perl-5.8.8

2017-09-18 Thread H.Merijn Brand
On Mon, 18 Sep 2017 17:57:39 +0100, Tim Bunce 
wrote:

> On Mon, Sep 18, 2017 at 11:46:34AM +0200, H.Merijn Brand wrote:
> > On Mon, 18 Sep 2017 11:19:08 +0200, "H.Merijn Brand"
> >  wrote:
> >   
> > > Makefile.PL states MIN_PERL_VERSION => '5.008', so I am still in the
> > > supported range.
> > > 
> > > My cause for digging is
> > > 
> > > Tie-Hash-DBD-0.16:
> > > - x86_64-linux-ld / 5.8.8:
> > >   - FAIL
> > > http://www.cpantesters.org/cpan/report/b06ed2c0-9b4b-11e7-b9b4-ee532540c4eb
> > > 
> > > When I wanted to make sure it is not already fixed by module
> > > updatesperl5.8.8 Makefile.PL
> > > 
> > > (and my requirements are too low), I noticed  
> 
> Looks like the failures relate to the recent the unicode changes.
> 
> > This is fixed by updating Storable, so the minimum requirement for
> > Storable's version is too low (or in our case, not required at all)
> > 
> > perl-5.8.8 with Storable 2.15 failed
> > perl-5.8.8 with Storable 2.51 passes
> > perl-5.8.9 with Storable 2.19 passes
> > 
> > So I'd suggest to make Storable version 2.19 a minimal requirement:  
> 
> Typo here, I think you meant 2.16 not 2.19.

I meant 2.19, as that is the first version after the failing 2.15 that
appeared in a stable perl. That implies that it is much harder to test
against 2.16 than it is to test against 2.19

Additionally, Storable 2.51 installs flawless on 5.8.8, I I made the
safer jump from 2.15 to 2.19 instead of to 2.16 which I did not test

If you agree with 2.19 because of this reasoning, I'll commit that
change with some comment

> > Agree?  
> 
> Yeap.
> 
> Thanks for looking after this!
> 
> Tim.

-- 
H.Merijn Brand  http://tux.nl   Perl Monger  http://amsterdam.pm.org/
using perl5.00307 .. 5.27   porting perl5 on HP-UX, AIX, and openSUSE
http://mirrors.develooper.com/hpux/http://www.test-smoke.org/
http://qa.perl.org   http://www.goldmark.org/jeff/stupid-disclaimers/


pgp33eYmesI1L.pgp
Description: OpenPGP digital signature


Re: DBI-git on perl-5.8.8

2017-09-18 Thread Tim Bunce
On Mon, Sep 18, 2017 at 11:46:34AM +0200, H.Merijn Brand wrote:
> On Mon, 18 Sep 2017 11:19:08 +0200, "H.Merijn Brand"
>  wrote:
> 
> > Makefile.PL states MIN_PERL_VERSION => '5.008', so I am still in the
> > supported range.
> > 
> > My cause for digging is
> > 
> > Tie-Hash-DBD-0.16:
> > - x86_64-linux-ld / 5.8.8:
> >   - FAIL
> > http://www.cpantesters.org/cpan/report/b06ed2c0-9b4b-11e7-b9b4-ee532540c4eb
> > 
> > When I wanted to make sure it is not already fixed by module
> > updatesperl5.8.8 Makefile.PL
> > 
> > (and my requirements are too low), I noticed

Looks like the failures relate to the recent the unicode changes.

> This is fixed by updating Storable, so the minimum requirement for
> Storable's version is too low (or in our case, not required at all)
> 
> perl-5.8.8 with Storable 2.15 failed
> perl-5.8.8 with Storable 2.51 passes
> perl-5.8.9 with Storable 2.19 passes
> 
> So I'd suggest to make Storable version 2.19 a minimal requirement:

Typo here, I think you meant 2.16 not 2.19.

> Agree?

Yeap.

Thanks for looking after this!

Tim.


Re: DBI-git on perl-5.8.8

2017-09-18 Thread H.Merijn Brand
On Mon, 18 Sep 2017 11:19:08 +0200, "H.Merijn Brand"
 wrote:

> Makefile.PL states MIN_PERL_VERSION => '5.008', so I am still in the
> supported range.
> 
> My cause for digging is
> 
> Tie-Hash-DBD-0.16:
> - x86_64-linux-ld / 5.8.8:
>   - FAIL
> http://www.cpantesters.org/cpan/report/b06ed2c0-9b4b-11e7-b9b4-ee532540c4eb
> 
> When I wanted to make sure it is not already fixed by module
> updatesperl5.8.8 Makefile.PL
> 
> (and my requirements are too low), I noticed

This is fixed by updating Storable, so the minimum requirement for
Storable's version is too low (or in our case, not required at all)

-- showing just stable perl versions and Storable version updates ...

$ corelist -a Storable
  v5.8.0 2.04
  v5.8.1 2.08
  v5.8.3 2.09
  v5.8.4 2.12
  v5.8.5 2.13
  v5.8.8 2.15
  v5.8.9 2.19
  v5.10.02.18
  v5.10.12.20
  v5.12.02.22
  v5.14.02.27
  v5.16.02.34
  v5.18.02.41
  v5.20.02.49
  v5.22.02.53
  v5.26.02.62

Working from there ...

perl-5.8.8 with Storable 2.15 failed
perl-5.8.8 with Storable 2.51 passes
perl-5.8.9 with Storable 2.19 passes

So I'd suggest to make Storable version 2.19 a minimal requirement:

--8<---
diff --git a/Makefile.PL b/Makefile.PL
index 4a58334..ee11d00 100644
--- a/Makefile.PL
+++ b/Makefile.PL
@@ -76,6 +76,7 @@ my %opts = (
},
 },
 PREREQ_PM => {
+ 'Storable' => 2.16,
  ( $^O eq 'MSWin32' ? ( 'File::Spec' => 3.31, ) : () ),
 },
 CONFLICTS => {
-->8---

Agree?

Reading Storable changelog, that makes sense, as the failing tests
involved dclone:

--8<---
Sat Mar 31 06:11:06 IST 2007   Abhijit Menon-Sen 

Version 2.16

1. Fixes to Storable::dclone, read_magic, retrieve_lscalar
2. Storable 0.1 compatibility
3. Miscellaneous compile/leak/test/portability fixes
-->8---

> $ perl5.8.8 Makefile.PL
> :
> :
> $ make
> :
> :
> $ make test
> :
> t/06attrs.t . 1/?
> #   Failed test 'Can set sth NAME'
> #   at t/06attrs.t line 263.
> 
> #   Failed test 'Can set sth NAME_lc'
> #   at t/06attrs.t line 263.
> 
> #   Failed test 'Can set sth NAME_uc'
> #   at t/06attrs.t line 263.
> 
> #   Failed test 'Can set sth NAME_hash'
> #   at t/06attrs.t line 263.
> 
> #   Failed test 'Can set sth NAME_lc_hash'
> #   at t/06attrs.t line 263.
> 
> #   Failed test 'Can set sth NAME_uc_hash'
> #   at t/06attrs.t line 263.
> # Looks like you failed 6 tests of 163.
> t/06attrs.t . Dubious, test returned 6 (wstat 1536, 0x600)
> Failed 6/163 subtests


-- 
H.Merijn Brand  http://tux.nl   Perl Monger  http://amsterdam.pm.org/
using perl5.00307 .. 5.27   porting perl5 on HP-UX, AIX, and openSUSE
http://mirrors.develooper.com/hpux/http://www.test-smoke.org/
http://qa.perl.org   http://www.goldmark.org/jeff/stupid-disclaimers/


pgpkclnMvqWfM.pgp
Description: OpenPGP digital signature