CGAL users discussion list

Hi David,

Thank you for all this. The first 2 patches have been already applied  
to the internal release and will be included in the next public release.

I was able to reproduce the problem in the 3rd item below with  
Exact_predicates_exact_constructions_kernel and even with arrangement  
(without history). We'll are looking into it.

Regarding your question, equal_2_object() should return true if and  
only if the graphs of the curves are the same. It's a good point, and  
we'll add it to the documentation.

Quoting David Keller 
<>:

> Hi,
>
> I have some issues with Arrangement_with_history_2:
>
>   1. operator== in Arr_with_history_2_functions.h misses a return value
>      for the case of non-self-assignments. Attached is a patch
>      (arr_with_hist_return_value.patch) which fixes the problem for me.
>   2. If VERBOSE is defined there are several issues with
>      function-name-lookups in derived objects leading to compile-time
>      errors (at least for gcc 4.0 and later). Attached is a patch
>      (sweep_line_debug.patch) to fix this problem.
>   3. Inserting an overlapping but "not equal" curve in an
>      Arrangement_with_history_2 leads to a runtime-error (precondition
>      failure in Arrangement_2_functions.h at line 1579). Attached is a
>      reproducer (arr_with_history_overlapping_curves.cpp). What should
>      traits::equal_2_object() implement here: Equal graph or equal
>      representation? The case may actually be rare. However, I'm not
>      quite sure if the current implementation gives the desired
>      behavior. If so, it would possibly be a good choice to mention
>      this in the documentation.
>
> Thanks and kind regards
> David Keller
> --
> You are currently subscribed to cgal-discuss.
> To unsubscribe or access the archives, go to
> https://lists-sop.inria.fr/wws/info/cgal-discuss