v1.2.0 release candidate

[haykh]

Structural Changes:

• Code of conduct added
• Add a list of explicitly supported (constantly upgraded) problem generators (Supported Problem Generators cleaned #92)

New Features:

• GR #83
• [REQUEST] Position-binned discretized phase-space output #48 (skipping this until future release)
• Checkpoint dumping (restart files) #67
• Add bulk velocity as moments to output #70
• alternative MPI communications for particles New Particle Send/Recv Routine #77
• field-downsampling in the output Downsampling of fields on output #73
• writing into separate files Output to separate files #75
• powerlaw injector Add a powerlaw injection to the archetypes #71
• Two-temperature Maxwellian (being tested in Supported Problem Generators cleaned #92)
• Option to disable GPU-aware-MPI (Optionally turn off GPU-aware MPI #105)
• Particle purge routine (being developed in Supported Problem Generators cleaned #92) (documented, not necessary to archetype)
• Box-reduced stats output (Stats output #106)
• Custom injector (being tested in Supported Problem Generators cleaned #92)

Improvements:

• New features and known bug fixes in checkpoints (Bugs + additional features in checkpoints #112)
• New zig-zag pusher with no explicit coordinate conversion (Zig-Zag Deposit + Cartesian Coordinate Push Without Explicit Coordinate Conversion #98)
• more consistent and general boundary condition setters (worked in magnetized shock #72)
• check for dx=dy=dz in Minkowski now with accuracy (fixed with 5cf30af)
• Log level controlled from input (ec7ff77 and b122d48)
• New structure of problem generators (-D pgen=... flag) (New PGen infranstructure #108)
• Small angle in QKs polar area (> SR and GR: bug in deposit: wrong jx3 in 2D #122)
• Symmetrized axis boundary setters for fields; increased range (> SR and GR: bug in deposit: wrong jx3 in 2D #122)
• Stabilized particle pusher and deposit near the axis in GR double precision (Dev/gr pusher axis #126)

Major bugs:

• [BUG] Matching boundaries act weirdly when enabled #114
• Linking error #62
• MPI error with mpich #60 (moved to v1.1.1)
• potential bug in field boundaries magnetized shock #72
• no need for explicit HDF5_ROOT in cmake [ERROR] Cmake Compiling Error for Loading Hdf5 Module for Output #82 (solved in Cleanup + Cosmetic changes #90)
• Errors in for certain MPI decompositions #74 (fixed in Fixes the checking of dx_min between domains #89)
• Added Taylor expansion to polar area calculation #110
• bug in 2D jz deposit (wrong component) > SR and GR: bug in deposit: wrong jx3 in 2D #122
• bug in jz GR deposit (missing alphain velocity) Dev/gr deposit #124 and fixed coordinate for deposit #125

Minor bugs:

• warning for template declaration bugfix for duplicate symbols #81
• field downsampling read both as scalar and as vector
• spectra output with MPI now works with adios2 format (not with hdf5) Spectra outout with MPI #119

Potentially breaking changes

• particle payloads are now a normal 2D Kokkos::View indexed as (prtl index, payload index)
• PGen::CustomFieldOutput signature changed (see example)
• problem generators are now picked by either their name -D pgen=streaming (for default ones), or -D pgen=pgens/khi (for the ones in entity-pgens repo), or via an absolute/relative path.
• by default, checkpoints are now written into <simname>.ckpt directory, unless otherwise specified in the input file