Repository navigation
Conversation
Co-Authored-By: Claude <noreply@anthropic.com>
Co-Authored-By: Claude <noreply@anthropic.com>
Co-Authored-By: Claude <noreply@anthropic.com>
Co-Authored-By: Claude <noreply@anthropic.com>
GiudGiud
left a comment
There was a problem hiding this comment.
considering you are at the page limit I think you just have a few substitutions to do and then it'd work for me
| \textbf{Meshing and I/O.} The \texttt{MeshTetInterface} abstract base class unifies libMesh's | ||
| interfaces to the bundled TetGen and Netgen tetrahedralizers, giving applications a single API for | ||
| generating volume meshes from boundary surfaces. Mesh I/O has expanded well beyond the 2006 | ||
| formats to include extended ExodusII/Nemesis, VTK, Abaqus, and NetCDF support. The element library | ||
| has added high-order and shell element types, along with generic element-quality metrics (aspect | ||
| ratio, minimum/maximum angle, and scaled Jacobian). | ||
|
|
||
| \textbf{Parallel infrastructure.} Parallel communication is now implemented in TIMPI, a required | ||
| dependency that began as libMesh-internal code and was split into its own independently versioned | ||
| and tested repository in 2019. TIMPI offers interchangeable strategies for push/pull-style parallel | ||
| data exchange, selected through the \texttt{SyncType} enumeration (\texttt{NBX}, | ||
| \texttt{ALLTOALL\_COUNTS}, \texttt{SENDRECEIVE}). The \texttt{NBX} strategy implements the | ||
| nonblocking dynamic sparse data exchange algorithm of Hoefler et al.~\cite{hoefler2010nbx}, which | ||
| avoids the all-to-all communication and a priori knowledge of message counts that earlier, simpler | ||
| schemes required. | ||
|
|
||
| \textbf{Numerics and solvers.} libMesh's generic numeric types (\texttt{TypeVector}, | ||
| \texttt{TypeTensor}, \texttt{VectorValue}, \texttt{TensorValue}, \texttt{DenseVector}, and |
There was a problem hiding this comment.
I would put these 3 first, with Ghosting right next to Parallel infrastruture
and put HDG last, as it's the most niche one
There was a problem hiding this comment.
I don't think that's true across research areas. A foundational HDG paper likely does not use anything other than a simple simplicial mesh nor would it employ parallelism. Someone who cares primarily about finite element theory would likely care little about these three topics. I think HDG is one of the more important topics for showing that we are an advanced finite element library. Both hybridization and static condensation are core finite element topics. These three are great for HPC and real physics applications for sure. Since I'm emphasizing libMesh's ability to support multiphysics, your re-ordering might be the right thing to do but it's not obvious to me. @roystgnr what do you think about topic ordering? You're going to be the primary author so your choice is the authoritative one
Co-Authored-By: Claude <noreply@anthropic.com>
Just now looking at the softwarex guidelines, AFAICT there's not a page limit (anymore?). There's a 4000 word count limit and the draft you reviewed @GiudGiud was only around 2200 words. So I'm going to take a swing at adding some more stuff. (Well really I'm probably going to have Claude take a first/second swing based off the research you already did and then I'll review) |
Co-Authored-By: Claude <noreply@anthropic.com>
Co-Authored-By: Claude <noreply@anthropic.com>
|
@roystgnr and @jwpeterson would appreciate your thoughts on any additional topics we'd want to cover, or topics you'd like to go into greater detail on. We're at about 2900 of our 4000 word limit now |
Opening a PR perhaps will help keep me working on this