Jose Luis Blanco-Claraco

Thanks, it would be great! In the meanwhile I tried with tricks redefining `__SIZE_TYPE__` and alike, but that's a dead end...

Sergey, this is definitively awesome, thanks! 👏👏 Sure, I'll try it and expand in a PR after testing... Cheers

PS: it would add a certain performance loss, but if you prefer it, I could also add an optional set of types loadable from the config file, *in addition to*...

@lyskov After your fix, all `cl.def(..)` cases are fixed regarding `size_t` and friends, that's all good. But now, the same problem remains in the auto-generated wrappers `PyCallBack_*`, see for example...

Maybe here? https://github.com/RosettaCommons/binder/blob/master/source/class.cpp#L701-L720

Thanks for the quick fix @lyskov ! I tested it and indeed it seems to have worked well. I'm waiting for packages to be built in armhf to fully confirm...

My package still have one remaining error for armhf but it's a too particular problem in my API... binder made the grade here too, thanks!! :100:

I found one more edge case: ``` void push_back(const double v) { internalPushBack(v); } void push_back(const std::string v) { internalPushBack(v); } void push_back(const uint64_t v) { internalPushBack(v); } ``` got...

For the records, I "solved" this in my project via a mass generation of wrapping code via binder, then applying hand-written patches to the problematic parts. Not ideal, but it's...

That makes a lot of sense, and I'm aware of the lot of dependencies implied by installing mrpt2... It's not been done yet because it's far from trivial, here upstream,...