I hope I'm using this mailing list thing right.

I know you said you looked in duckmatic, but this bit of code looks right
to me:
https://github.com/synfig/synfig/blob/master/synfig-studio/src/gui/duckmatic.cpp#L2653
I don't know much about how ducks work, though, so maybe that's not what
you're looking for.

Looking at the stack trace, however, I think the source of the bug might be
here:
https://github.com/synfig/synfig/blob/master/synfig-studio/src/synfigapp/actions/valuedesclink.cpp#L124

It's trying to check if the input ValueDesc has a sub_name, but if it
doesn't have a parent ValueDesc, it won't have a sub_name. So changing the
condition to this...
value_desc.parent_is_value_desc() && !value_desc.get_sub_name().empty()
...may work.

Just my two cents.


On Fri, Sep 4, 2015 at 8:26 AM, <d.j....@free.fr> wrote:

> Hello,
>
> I'm looking at fixing :
> #889 - Crash on linking regular vertice and width point center TYPE_WIDTH
> http://www.synfig.org/issues/thebuggenie/synfig/issues/889
>
> the crash come the assert(parent_is_value_desc()) from here :
>
> https://github.com/synfig/synfig/blob/master/synfig-studio/src/synfigapp/value_desc.h#L340
>
> inside this call stack :
>
>         __assert_fail() at 0xb56a57d7
>         synfigapp::ValueDesc::get_sub_name() at value_desc.h:340 0x86e6a5f
>         synfigapp::Action::ValueDescLink::set_param() at
> valuedesclink.cpp:124 0xb7d8ff1c
>         synfigapp::Action::Base::set_param_list() at action.cpp:377
> 0xb7e1c1a5
>         studio::Instance::process_action() at instance.cpp:1 077 0x88952f4
>         sigc::bound_mem_functor2<void, studio::Instance, std::string,
> synfigapp::Action::ParamList>::operator() at mem_fun.h:1 917 0x88a5f13
>         <...more frames...>
>
> The width point is treated has a composite value_node (but should'nt maybe
> ?) and i can't figure where/when width point are effectively created...
>
>
> 1) Width point duck .... when and where are they created ? (i have checked
> inside state_xxxx , and duckmatic ... )
> 2) Maybe Width points shouldn't be treated has compisite valuenode .... ?
>
>
> see'ya!
>
>
> ------------------------------------------------------------------------------
> _______________________________________________
> Synfig-devl mailing list
> Synfig-devl@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/synfig-devl
>
------------------------------------------------------------------------------
_______________________________________________
Synfig-devl mailing list
Synfig-devl@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/synfig-devl

Reply via email to