Alex Margolin

@maumueller Yes, there are some open technical issues, and that's why I've submitted it as a "draft PR". To clarify, this runs on my machines, but the best way to...

@ggouaillardet @rhc54 I probably didn't have the same intention as @cponder , but I stumbled across what seems to be the same issue: `mpirun --rank-by core XYZ` just fails (OMPI...

@shamisp afraid not, and I don't expect it'll be resolved soon. I'm posting stuff I plan to upstream once it does - I just don't seem to have permissions to...

> how is it different from using regular khash with a custom key type that contains multiple values? This is the result of some research, and the paper is still...

> IIUC, the expected lookup performance of the special multi-dim implementation should be better than default khash with a vector key? No, I'm afraid the lookup is not faster, just...

> @alex--m Can you give a bit more details how it will be used ? AKA how we will benefit from nearest neighbor lookup speedup. Sure. The basic Idea is...

@dmitrygx Yes, make sense to me. It's been a long time since I wrote this timerq code, so can't be 100% sure the tests fit the API changes, but I...

waiting for #5609

In addition to the more specific issue #19, I'm sure every collective component would benefit from topology information. For reference, here's what UCG expects from MPI today: ```` /** *...

@adammoody Yeah, I built with an older GCC for now, so I don't mind waiting. Thank you for including a fix into the roadmap!