Aliaksandr Yakutovich

I am trying to do `git checkout` following the suggestions described above. It seems to me that those two steps are still not enough: ``` r.reset_index(r[b"refs/heads/branchname"].tree) r.refs.set_symbolic_ref(b"HEAD", b"refs/heads/branchname") ``` This...

> Although it is good to have the possibility to define this on a `Code` instance, should there also not be some kind of enforcement in AiiDA? I understand that...

> That suggests that there _is_ any interaction between the two and the metadata option is leading. But I don't see where in `aiida-core` the `Code.get_withmpi` is used at all....

> My first question is whether this should be a flag? I guess it should be a flag. Or, in other words, I don't understand why it shouldn't. > It...

I finished the implementation. @sphuber and @unkcpz can you please have a look before I fix the tests? Another thing: I would prefer to rename `withmpi` option to `with_mpi` are...

Ok, I tested it on the `provenance_redesign` branch and looks like the trajectory object can't be shown there either. The error message is different though: ``` Traceback (most recent call...

Just to let you know that the problem is still there for 2.7.7, unless you consider adding the code below as the fix ```python def on_change(change): index = change['new'] if...

This may help. If I do `view.n_components = 1` before displaying the widget for the first time - it works without a problem.

Thanks, @hainm, with setting attribute it works. I was just wondering: is it expected behaviour? I had the impression that nglview's interface is very much in line with jupyter widgets,...

Alright, thanks for the explanation. Nice package by the way!