I've been annoyed by this behavior as well.  I had
planned on adding a permanent work-around for the
lvalue subs bug by automatically inserting the
'(my $tmp = $pdl->lvalue()) .=' into code when
compiled for the debugger.  However, that still
leaves the "step into" problem.

Maybe there is a way to customize the debug/compile
process so that n will skip "through" the lvalue sub but
s would go into it...

Maybe someone on perlmonks might have an idea
for a fix....

--Chris

On Fri, Aug 5, 2011 at 11:46 AM, Doug Hunt <[email protected]> wrote:
> Hi all:  I would like to see the annoying problem with debugging lvalue PDL
> routines fixed.
>
> Currently, when debugging a PDL code, I am forced to step into any lvalue
> routine (such as 'slice') and step through the entire routine before
> returning to my code.  This is at root a perl problem, but the eventual
> solution may involve perl and PDL changes.
>
> If I have time I would like to look into this.
>
> --Doug
>
> [email protected]
> Software Engineer
> UCAR - COSMIC, Tel. (303) 497-2611
>
> On Fri, 5 Aug 2011, Chris Marshall wrote:
>
>> Hope that a better subject generates better
>> feedback....
>>
>> --Chris
>>
>> On Wed, Jul 27, 2011 at 2:11 PM, Chris Marshall <[email protected]>
>> wrote:
>>>
>>> Executive Summary:
>>> * Let's have a PDL-2.4.10 release this Fall.
>>> * What would you like to see in it?
>>> * Can you help?
>>>
>>> Long Version:
>>>
>>>     The current PDL-2.4.9_003 release is testing
>>> well with CPAN Testers and with the new pthreads
>>> support,  I would like to see a PDL-2.4.10 official
>>> release in the next 2-3 months.
>>>
>>> I would be happy for that release to include bug
>>> fixes for the catastrophic bugs (e.g., segfaults!)
>>> and the "easy-fixes" bugs (i.e., easy fixes for the
>>> right developer).
>>>
>>> Among my own development projects, I would
>>> like to complete the NiceSlice2 implementation
>>> and finish back compatibility for the pdl2
>>> implementation (so we can go back to a single
>>> PDL shell, even if the original perldl one is being
>>> called behind the scenes).
>>>
>>> What would you like to see for PDL-2.4.10?
>>>
>>> --Chris
>>>
>>
>> _______________________________________________
>> Perldl mailing list
>> [email protected]
>> http://mailman.jach.hawaii.edu/mailman/listinfo/perldl
>

_______________________________________________
Perldl mailing list
[email protected]
http://mailman.jach.hawaii.edu/mailman/listinfo/perldl

Reply via email to