Edoardo Zoni

Hi @dpgrote, do you have any feedback on this PR and/or the last comment https://github.com/BLAST-WarpX/warpx/pull/6053#issuecomment-3255569948 in particular?

> As for the big checksum change, I think the difference starting from roundoff can get amplified a lot because the solver is iterative. I think part of what seems...

@oshapoval There are some conflicts that need to be resolved, otherwise the pre-commit CI check cannot run properly.

@RemiLehe Since I was in the meeting where this was discussed, I feel like I can ask a question. Is this basically what you have in mind (forgetting the vector...

I'll let @dantargz or @ax3l answer @titoiride's question: > I wasn't aware of this when I opened https://github.com/BLAST-WarpX/warpx/issues/6257. I'm curious, how do you handle the case with moving window then?...

I have no objections to the fix per se, but isn't this (a tridiagonal matrix algorithm) something basic enough to belong better to a numerical library like AMReX rather than...

@JustinRayAngus Some info on the CI configuration topic. This is the step where we configure and compile WarpX before running the CI tests in the Azure pipelines: https://github.com/BLAST-WarpX/warpx/blob/83cfdd028a5aba612f1d03cb0cf22c432f0f98b8/.azure-pipelines.yml#L142-L158 You see...