Richard Gowers

@Luthaf fwiw, we had a GSOC student last year looking at domain decomposition methods for limited distance searches (ie distances up to a threshold, here bond lengths). The full details...

It would be nice for performance reason to be single precision. (Hugo and I have work on vectorising where width matters) Also the current pattern of inputting single and returning...

@yuxuanzhuang thanks for adding the test - it looks like it found that h5md format has a custom serialisation mechanism

Makes more sense to remove from `run()` rather than `__init__` imo

Yeah the lines between prepare and init are very blurred. I think one differentiator could be parallelism, which iirc does a prepare on each worker which then gets merged, whereas...

Yeah it’s not actually well defined admittedly, this is my memory of how pmda works. I think one of the main goals of the structure of AnalysisBase is to make...

@mdaviesNCL hmm odd, that is usually the error message for when upstream has had additions. Worst case, you can rename your local branch to something "safe", re-check out this remote...

@mdaviesNCL are you still having trouble pushing up changes?

So currently if you've loaded a PDB file, then any `REMARK` lines are accessible at `Universe.trajectory.remarks`, but they aren't parsed into anything better than the raw lines from the file....

Ok I see, its possible there are object topologyattrs that aren’t string, so hacking the internet mixin seems good yeah. On Sat, Apr 23, 2022 at 09:19, Lily Wang ***@***.***>...