[yt-dev] lost weak reference

[david collins]

> If you still have the trace back, could you file a bug? I'd like to fix it
> so that if this does happen again, reconnection works - sorry it didn't this
> time, I know how much it stinks to lose wprk/time like that.


I don't have that trace back specifically, but I re-created it with a
dummy script and submitted a ticket.   While doing this I realized
that it might be because the _hierarchy_ was also a weakref, so I
probably needed to replace that, as well (?)

Thanks for your input. The real moral of the story is "dave should be
careful with pointers"

d.

>
> Matt
>
> On Aug 10, 2012 12:56 PM, "david collins" <antpuncher at gmail.com> wrote:
>>
>> > That is one way to fix this, yes, although maybe making sure the
>> > reference doesn't go out of scope would be easier?
>>
>> I'll definitely make sure that it doesn't croak in the future (I
>> believe I've sorted it out since I sent this email-- I did, in fact,
>> do something dumb) but this bit took a while to run so if I _can_ fix
>> the thing I have that'd be sweet.  I did try re-attaching the
>> parameter file, but it didn't seem to take-- I think I'll just punt
>> and read papers today and re-run this analysis without the dumb-ness.
>>
>> Thanks a ton!
>> d.
>>
>> >
>> > On Fri, Aug 10, 2012 at 12:39 PM, david collins <antpuncher at gmail.com>
>> > wrote:
>> >> Ah.  I know what the parameter file _should_ be, can I go re-assign it
>> >> to the clumps?
>> >>
>> >> d.
>> >>
>> >> On Fri, Aug 10, 2012 at 10:33 AM, Matthew Turk <matthewturk at gmail.com>
>> >> wrote:
>> >>> Hi Dave,
>> >>>
>> >>> Looks like the reference is not to the data object, but the parameter
>> >>> file.  From this context it's tough to see how that reference expired,
>> >>> though.  It could have happened if something like this were done:
>> >>>
>> >>> data = load("something").h.all_data()
>> >>>
>> >>> or if you ha a pf variable that was not held anywhere else that then
>> >>> went out of scope.
>> >>>
>> >>> -Matt
>> >>>
>> >>> On Fri, Aug 10, 2012 at 12:31 PM, david collins <antpuncher at gmail.com>
>> >>> wrote:
>> >>>> Hi, all--
>> >>>>
>> >>>> Is there a way to reconstruct a bad data object?  I'm making a bunch
>> >>>> of clumps, but some how (unsure how right now)  the weak reference to
>> >>>> "data" on  some of them is evaporating.
>> >>>>
>> >>>> Thanks,
>> >>>> d.
>> >>>>
>> >>>>>>> the_clumps[1].data
>> >>>> Traceback (most recent call last):
>> >>>>   File "<stdin>", line 1, in <module>
>> >>>>   File
>> >>>> "/Users/dccollins/local2/src/yt-hg/yt/data_objects/data_containers.py",
>> >>>> line 3006, in __repr__
>> >>>>     s = "%s (%s): " % (self.__class__.__name__, self.pf)
>> >>>> ReferenceError: weakly-referenced object no longer exists
>> >>>>
>> >>>>
>> >>>>
>> >>>> --
>> >>>> Sent from my computer.
>> >>>> _______________________________________________
>> >>>> yt-dev mailing list
>> >>>> yt-dev at lists.spacepope.org
>> >>>> http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>> >>> _______________________________________________
>> >>> yt-dev mailing list
>> >>> yt-dev at lists.spacepope.org
>> >>> http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>> >>
>> >>
>> >>
>> >> --
>> >> Sent from my computer.
>> >> _______________________________________________
>> >> yt-dev mailing list
>> >> yt-dev at lists.spacepope.org
>> >> http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>> > _______________________________________________
>> > yt-dev mailing list
>> > yt-dev at lists.spacepope.org
>> > http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>>
>>
>>
>> --
>> Sent from my computer.
>> _______________________________________________
>> yt-dev mailing list
>> yt-dev at lists.spacepope.org
>> http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>
>
> _______________________________________________
> yt-dev mailing list
> yt-dev at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>



-- 
Sent from my computer.