Proposed changes for 5.4.0

[rjleveque]

There are some changes that should be released soon as 5.3.1.

There are several other changes in the works that change calling sequences or Makefiles and/or can lead to different grid patches being generated in amrclaw. A partial list is below. Other things and/or discussion?

• fixes to filpatch and CFL computation with dimensional splitting. These are included in Faster bound amrclaw#149.
• bc2amr calling sequence wrong in amrclaw routines geoclaw#171 should not be merged. Instead we plan to change the bc routines in amrclaw and geoclaw to remove more things from the calling sequences.
• Faster bound amrclaw#149 and Faster bound geoclaw#173 improve the way ghost cells are filled, can be much faster,
• Check_twofiles.f and run_with_restart.py to automatically restart interrupted run. amrclaw#144 gives a way to alternate between two checkpoint file names rather than accumulating lots of checkpoint files.
• modular makefiles clawutil#88 suggests a new way to simplify Makefiles, by having a default list for each source library that would be modified in application directories, rather than listing all source files. The discussion on that PR suggests reconsidering libraries but that might be put off to 6.0. More discussion welcome, but at least this PR is backward compatible.
• [WIP] Created regions_module for 3d. Only basic functionality (read/write regions.data) at this point. amrclaw#131 is a start at regions in 3d amrclaw. Include? (put off for later)
• Add pressure term to Riemann solver
• Rework region specification (still some work to do in the future)
• Add basic parallelism to matplotlib (done in Parallel plotting visclaw#174)

[mandli]

What about considering a rebuilt region specification in general? Merge say the dtopo, topo, qinit, and general refinement specifcation?

[rjleveque]

I'm all in favor of that, along with adding regions in 3d and eliminating the multiple checking of regions in geoclaw (see clawpack/geoclaw#117), but I'm not sure we can get that done for 5.4.

[mandli]

Not sure either. This would be a good thing to discuss perhaps with Dave George as he had some ideas along these lines. At least we could outline a plan and what needs to be done.

[ketch]

We should probably update (and hopefully close) this issue since we have a 5.4.0 prerelease out.

[mandli]

+1

[ketch]

Most of the linked pull requests have been merged, or they have been superseded by another that has been merged. However, I'm not very familiar with them so I'll leave it to those who are (mostly @rjleveque) to check boxes.

[cjvogl]

As for the regions in 3d item, I do have code from the seismic work that implements regions in 3d through flag2refine.f... however, this deviates from how regions are now specified in 2d. Just let me know if it should be included "as is" and I will open a PR.

[mandli]

@cjvogl I would encourage you to open a PR for the regions and we can see what can be merged.

[cjvogl]

Relevant PR is now open for discussion at clawpack/amrclaw#187.