Re: [perl #43334] [TODO] config/auto/icu.pm: Write unit tests

2008-07-07 Thread chromatic
On Monday 07 July 2008 17:37:41 Jason Cole wrote: > James Keenan wrote: > > (Coleoid: Does this update cause any problems for you on Win32 without > > ICU?) > > No visible problems from prove and Configure. Output seems the same and > indicates success. Pasted in #parrot. > > > Trying to 'make'

Re: [perl #43334] [TODO] config/auto/icu.pm: Write unit tests

2008-07-07 Thread Jason Cole
James Keenan wrote: > (Coleoid: Does this update cause any problems for you on Win32 without > ICU?) No visible problems from prove and Configure. Output seems the same and indicates success. Pasted in #parrot. Trying to 'make', though, I've started emitting hundreds of 'warning: control re

Re: [perl #43334] [TODO] config/auto/icu.pm: Write unit tests

2008-07-07 Thread François Perrad
James Keenan via RT a écrit : On Mon Jul 07 04:46:52 2008, [EMAIL PROTECTED] wrote: Do you think you could 'svn up', reconfigure, and then let me know if your patch is still needed? Also send output of 'prove -v t/steps/auto_icu*.t' and 'perl Configure.pl --verbose-step=auto::icu'. Thanks.

Re: [perl #43334] [TODO] config/auto/icu.pm: Write unit tests

2008-07-07 Thread François Perrad
James Keenan via RT a écrit : Please review the patch attached. Note the following: 1. As mentioned in my last post in this RT, the flow in this step class's runstep() method is quite convoluted. I tried to improve it, but this step still has five different points at which it can return. I h

Re: [perl #43334] [TODO] config/auto/icu.pm: Write unit tests

2008-07-01 Thread Will Coleda
-- Will "Coke" Coleda On Jul 1, 2008, at 23:01, "James Keenan via RT" <[EMAIL PROTECTED] > wrote: On Tue Jul 01 18:34:25 2008, coke wrote: I would err on the side of removing them. No point in keeping unused items after the refactor, especially if you're going to end up having to write te

Re: [perl #43334] [TODO] config/auto/icu.pm: Write unit tests

2008-07-01 Thread Will Coleda
On Tue, Jul 1, 2008 at 8:10 PM, James Keenan via RT <[EMAIL PROTECTED]> wrote: > Please review the patch attached. Note the following: > > 1. As mentioned in my last post in this RT, the flow in this step > class's runstep() method is quite convoluted. I tried to improve it, > but this step stil