Thomas Madlener

Thanks for the clarification. It looks like there is no really solution in this case. Given that the recommended way in c++ to do the interface is to use `unique_ptr`...

I am not entirely sure what exactly triggered this as I rebuilt the complete stack so there might be other dependencies that changed as well. Another possible reason could be...

Nothing more clever from my side, I am afraid. One other possibility would be to not go for the absolute minimum requirements here, but rather find a somewhat inclusive set...

I am not sure how easy it is to pick this up, but @wlav seems to have been able to fix this with this commit: https://github.com/wlav/CPyCppyy/commit/02072bbd818501364b37c133364e02976af4cc10

I think in this case the `cppyy` that we pick up is the one that comes bundled with ROOT[^1], and this seems to be working in all other OS'es (or...

Just tested this and we get the same error message: ``` Traceback (most recent call last): File "", line 1, in AttributeError: has no attribute 'type'. Full details: type object...

After just running into this (with `meson` in my case); Is there any news on this or on #31939? Or is this currently waiting for the upstream situation to resolve...

If putting the datasource into EDM4hep would be (more or less) easily possible, I would also be in favor of having it there. That should make maintenance a bit easier...

> There is a lot of analysis code which uses the "old way" and also the performance of the datasource way needs to be better understood. Good point. Didn't think...

Also the CMake config needs adaptation most likely: https://github.com/HEP-FCC/FCCAnalyses/blob/8ce567fdf79e4a78d0f389dfa984aeabb596c154/analyzers/dataframe/CMakeLists.txt#L42 Alternatively if these are private headers, they should be declared as such (and removed from the public headers)