rolanddenis

Can one of you test to replace `std::max_element` by `boost::first_max_element` that is available in `boost/algorithm/minmax_element.hpp`? It is a proposed fix from https://github.com/boostorg/graph/issues/175 even if it will not solve the fact...

Thx, agree for the PR ! BTW I was more thinking in replacing by the boost version **only** where needed (with smart iterators like `ConstIteratorAdaptor`), so probably only in: -...

I get these warnings only if I enable `WITH_QGLVIEWER` (and also for older versions of GCC, 7 & 8)... same for you?

I should have checked the code before: these pragma are parsed only if `WITH_QGLVIEWER` is set: https://github.com/DGtal-team/DGtal/blob/master/src/DGtal/base/Common.h#L45 :sweat_smile:

Clang is supposed to support the `#pragma GCC` but I don't know if XCode behaves differently... @dcoeurjo can you check on your side after removing the `#pragma clang`?

> Yes! I have `WITH_QGLVIEWER` enabled. I should re-check if I get these in gcc10. Same warnings in GCC 10.1 (BTW thanks for the information that v10 was released :wink:...

Still fails on my side with GCC 10.2.0, Eigen 3.3.9 and Boost 1.75.0 ... Does it work on your side ?

Fails also with LLVM 10.0.0, Boost 1.75.0 and Eigen 3.3.9 (Ubuntu), strange...

I cannot test on Boost 1.58.0 anymore but looking at code of `DiscreteExteriorCalcul` shows that it depends on Boost for `boost::array` and `boost::unordered_map`. Naively replacing `boost::unordered_map` by `std::map` to define...

In [`DiscreteExteriorCalculus::updateIndexes`](https://github.com/DGtal-team/DGtal/blob/master/src/DGtal/dec/DiscreteExteriorCalculus.ih#L609), index associated to each cell depends on the read order of `myCellProperties`. In [`DiscreteExteriorCalculus::derivative`](https://github.com/DGtal-team/DGtal/blob/master/src/DGtal/dec/DiscreteExteriorCalculus.ih#L324), this index is used as column index to fill the resulting sparse array, thus...