Re: [boost] [Review] We're halfway through the GGL review

Subject: Re: [boost] [Review] We're halfway through the GGL review
From: Barend Gehrels (barend_at_[hidden])
Date: 2009-11-15 12:53:37

Hi Fabio,

Thanks for your interest and for reviewing our library!

>
> polygon p1 = ...;
> polygon p2 = ...;
>
> polygon_set ps1 = make_union(p1, p2);
> polygon_set ps2;
> intersection(ps1, p2, ps2);
The polygon_set is implemented as a multi-polygon within GGL (and indeed 
you can call it polygon_set, if it fulfils the concept). We didn't merge 
the Multi* concepts into the generic algorithms, in other words, we 
don't require anyone to use a multi* (of course algorithms can be used 
for multi*).

Therefore, union and intersection work with output iterators. So:
union_inserter(p1, p2, std::back_inserter(ps1));

It is possible that there will be a "intersection" and "union_" as well, 
detecting if ps1 is a multi-polygon and automatically inserting. But 
that is more a sort of synonym, similar way to do things.

The make_union would require an r-value and if C++0x is there, it can be 
implemented without copying the whole resultset. Indeed, it fits into 
the design and naming conventions.

>
> On the whole I am very pleased with the documentation, its easy to 
> read and reasonably complete.
> I have a general problem with doxygen generated docu with regard to 
> generic libraries.[...]
I understand. Several people have addressed this so we will certainly 
improve the documentation. Thanks for your structure proposal.

Regards, Barend