Releases: GEOS-ESM/MAPL
MAPL 2.64.1
This is a patch release of MAPL 2.64. It restores a workaround for GNU for binary restart reads. This is used in the GEOS single column model (SCM).
This is zero-diff to v2.64.0
This version of MAPL was tested with:
- Intel Fortran Classic (
ifort) 2021.13 - Intel Fortran (
ifx) 2025.3 - GCC 13.2.0, 14.2.0, and 15.2.0
- NAG 7.2
The libraries this version of MAPL is currently tested with are below.
- Baselibs 8.20.0
- netcdf-c 4.9.2
- netcdf-fortran 4.6.1
- ESMF 9.0.0b03 (Note: MAPL only requires 8.6.1 at the moment)
- GFE 1.20.0
- gFTL 1.16.0
- gFTL-shared 1.11.0
- pFUnit 4.13.0 (optional)
- fArgParse 1.10.0 (if
-DBUILD_WITH_FARGPARSE=YES, default=YES) - pFlogger 1.17.0 (if
-DBUILD_WITH_PFLOGGER=YES, default=YES)
- UDUNITS2 2.28.8
Also, if you build with Baselibs and/or -DUSE_F2PY=ON, you should use (at least):
- ESMA_cmake v3.68.0
We recommend most external users set -DUSE_F2PY=OFF
We also require CMake 3.24 or higher.
From CHANGELOG.md
Fixed
- Restore workaround for binary restart reads with GNU
- Brings back changes from PR #1038: Added a workaround for a gfortran bug that handles end-of-file incorrectly (returns IOSTAT=5001).
What's Changed
Full Changelog: v2.64.0...v2.64.1
Contributors
Assets 3
MAPL 2.64.0
This release of MAPL has an update for MAPL_LocStream to support work ongoing in GEOSldas, adding the ability to pass Pfaffstetter index.
All tests show this is zero-diff to MAPL 2.63
This version of MAPL was tested with:
- Intel Fortran Classic (
ifort) 2021.13 - Intel Fortran (
ifx) 2025.3 - GCC 13.2.0, 14.2.0, and 15.2.0
- NAG 7.2
The libraries this version of MAPL is currently tested with are below.
- Baselibs 8.20.0
- netcdf-c 4.9.2
- netcdf-fortran 4.6.1
- ESMF 9.0.0b03 (Note: MAPL only requires 8.6.1 at the moment)
- GFE 1.20.0
- gFTL 1.16.0
- gFTL-shared 1.11.0
- pFUnit 4.13.0 (optional)
- fArgParse 1.10.0 (if
-DBUILD_WITH_FARGPARSE=YES, default=YES) - pFlogger 1.17.0 (if
-DBUILD_WITH_PFLOGGER=YES, default=YES)
- UDUNITS2 2.28.8
Also, if you build with Baselibs and/or -DUSE_F2PY=ON, you should use (at least):
- ESMA_cmake v3.68.0
We recommend most external users set -DUSE_F2PY=OFF
We also require CMake 3.24 or higher.
From CHANGELOG.md
Added
- Added
pfaf_indextoMAPL_Locstream - Added a simple MAPL_LocstreamCreate
Changed
- Enforced presence of
gridwhen reading binary tile file
What's Changed
- added pfaf index to MAPL_Locstream by @weiyuan-jiang in #4000
- GitFlow: Merge develop into main for 2.64 release by @mathomp4 in #4229
Full Changelog: v2.63.1...v2.64.0
Contributors
Assets 3
MAPL 2.63.1
This is a patch release on MAPL 2.63. This release fixes an issue needed for 1-d tile space output in netCDF format.
All tests show it zero-diff to MAPL 2.63.0
This version of MAPL was tested with:
- Intel Fortran Classic (
ifort) 2021.6.0 and 2021.13 - Intel Fortran (
ifx) 2025.3 - GCC 13.2.0, 14.2.0, and 15.2.0
- NAG 7.2
The libraries this version of MAPL is currently tested with are below.
- Baselibs 8.20.0
- netcdf-c 4.9.2
- netcdf-fortran 4.6.1
- ESMF 9.0.0b03 (Note: MAPL only requires 8.6.1 at the moment)
- GFE 1.20.0
- gFTL 1.16.0
- gFTL-shared 1.11.0
- pFUnit 4.13.0 (optional)
- fArgParse 1.10.0 (if
-DBUILD_WITH_FARGPARSE=YES, default=YES) - pFlogger 1.17.0 (if
-DBUILD_WITH_PFLOGGER=YES, default=YES)
- UDUNITS2 2.28.8
Also, if you build with Baselibs and/or -DUSE_F2PY=ON, you should use (at least):
- ESMA_cmake v3.68.0
We recommend most external users set -DUSE_F2PY=OFF
We also require CMake 3.24 or higher.
From CHANGELOG.md
Changed
- Combine MAPL_Grid_interior and MAPL_GridgetInterior
- This is needed for 1-d tile space output in netCDF4 format
What's Changed
- Bump actions/checkout from 5 to 6 by @dependabot[bot] in #4195
- Feature/wjiang/combine interio by @weiyuan-jiang in #4132
- GitFlow: Merge develop into main for 2.63.1 release by @mathomp4 in #4205
Full Changelog: v2.63.0...v2.63.1
Contributors
Assets 3
MAPL 2.63.0
This release of MAPL 2 has many new updates and fixes.
The main updates are to handling of Tile grids for use in GEOSldas. There are also updates to the ACG Writer for our ongoing transition to MAPL3 as well as many internal changes to the CI.
All testing with GEOSgcm shows this to be zero-diff to MAPL 2.62
This version of MAPL was tested with:
- Intel Fortran Classic (
ifort) 2021.6.0 and 2021.13 - Intel Fortran (
ifx) 2025.3 - GCC 13.2.0, 14.2.0, and 15.2.0
- NAG 7.2
The libraries this version of MAPL is currently tested with are below.
- Baselibs 8.20.0
- netcdf-c 4.9.2
- netcdf-fortran 4.6.1
- ESMF 9.0.0b03 (Note: MAPL only requires 8.6.1 at the moment)
- GFE 1.20.0
- gFTL 1.16.0
- gFTL-shared 1.11.0
- pFUnit 4.13.0 (optional)
- fArgParse 1.10.0 (if
-DBUILD_WITH_FARGPARSE=YES, default=YES) - pFlogger 1.17.0 (if
-DBUILD_WITH_PFLOGGER=YES, default=YES)
- UDUNITS2 2.28.8
Also, if you build with Baselibs and/or -DUSE_F2PY=ON, you should use (at least):
- ESMA_cmake v3.68.0
We recommend most external users set -DUSE_F2PY=OFF
We also require CMake 3.24 or higher.
From CHANGELOG.md
Fixed
- Removed obsolete pfio test
- Fix a memory leak in trajectory sampler due to misuage of FieldRegrid which should be FieldRedist
- Fixed
mapl_acg.cmaketo allow for more than one StateSpecs file per target - Fix NVHPC issue with IEEE halting code
- Fix a misspelled
MAPL_LIBRARY_TYPEinvertical - Workaround for Flang
Added
- Added updated version of
MAPL_GridCompSpecs_ACG_writer.py - Added 'Gridname' attribute to history output
- Create TilgridIO's outbundle from output grid and deallocate mGriddedIO in History
- Added TileGridIO.F90 to output NC4 History file in tile space. The collection's format should be 'CFIO'
- Added bit shave to TileGridIO
Changed
- Update CI
- Use Baselibs 8.20.0
- Add gcc15 test
- Use organization reusable workflows
- Run CI build test with NAG on PRs
- Update
README_ACG_WRITER.md - Update
components.yamlESMA_envv5.16.0- Update to Baselibs 8.20.0
- GFE v1.20.0
- gFTL v1.16.0
- gFTL-shared v1.11.0
- fArgParse v1.10.0
- pFUnit v4.13.0
- yaFyaml v1.6.0
- pFlogger v1.17.0
- NCO 5.3.5
- This is mainly for ifx 2025.2 support
- Requires CMake 3.24
- GFE v1.20.0
- Update to Baselibs 8.20.0
ESMA_cmakev3.68.0- Update
ifxflags to match as close as possible toifortflags - Update site detection for NAS
- Updates for f2py and flang
- Update
Removed
- Removed TileIO.F90. It is integrated into TileGridIO.F90
What's Changed
- v2: Update CI, Add gcc15 test by @mathomp4 in #4094
- v2: Add support for more than one StateSpecs per target by @mathomp4 in #4103
- Auto GitFlow - main → develop by @github-actions[bot] in #4107
- v2: Update CI to use reusable workflows by @mathomp4 in #4111
- v2: Fix issue with NVHPC and halting by @mathomp4 in #4110
- NC4 file in tile space for History output by @weiyuan-jiang in #4097
- Auto GitFlow - main → develop by @github-actions[bot] in #4120
- Auto GitFlow - main → develop by @github-actions[bot] in #4123
- Fix a memory leak in trajectory sampler by @metdyn in #4057
- Feature/wjiang/destroy mgridded by @weiyuan-jiang in #4126
- Bump devops-infra/action-pull-request from 0.6.1 to 1.0.0 by @dependabot[bot] in #4130
- v2: Update CI to Baselibs 8.20 by @mathomp4 in #4137
- v2: Fix misspelled MAPL_LIBRARY_TYPE by @mathomp4 in #4145
- Bump devops-infra/action-pull-request from 1.0.0 to 1.0.2 by @dependabot[bot] in #4140
- Feature/wjiang/gridname in history by @weiyuan-jiang in #4148
- Bump actions/upload-artifact from 4 to 5 by @dependabot[bot] in #4153
- Document ACG Writer by @darianboggs in #4157
- Auto GitFlow - main → develop by @github-actions[bot] in #4163
- remove obsolete pfio test by @weiyuan-jiang in #4139
- v2: Add aider dot directories to gitignore by @mathomp4 in #4141
- v2: Update to ESMA_env 5.16.0, ESMA_cmake 3.68.0 by @mathomp4 in #4173
- Fixes #4172 - flang workarounds by @tclune in #4174
- v2: Turn on NAG build CI on PRs by @mathomp4 in #4181
- make bit shaving work on nc4 tile space output by @weiyuan-jiang in #4190
- v2: Prepare for 2.63.0 release by @mathomp4 in #4193
- GitFlow: Merge develop into main for 2.63.0 release by @mathomp4 in #4196
Full Changelog: v2.62.2...v2.63.0
Contributors
Assets 3
MAPL 2.62.2
This is a patch release of MAPL 2.62. This fixes a bug when one of the two names in an ExtData2G rule for a vector is contained is a substring in another item, for example UA;VA and EVAP
Testing shows this is zero-diff to MAPL 2.62.1
This version of MAPL was tested with:
- Intel Fortran Classic (
ifort) 2021.6.0 and 2021.13 - Intel Fortran (
ifx) 2025.2- NOTE:
ifx2025.2 support is currently a bit of a work-in-progress as it requires changes in upstream GFE libraries (pFlogger and pFUnit) as well as ESMA_cmake v3.65.0. Once releases are made of the GFE libraries, this will become simpler.
- NOTE:
- GCC 13.2.0, 14.2.0, and 15.2.0
- NAG 7.2
The libraries this version of MAPL is currently tested with are below.
- Baselibs 8.19.0
- netcdf-c 4.9.2
- netcdf-fortran 4.6.1
- ESMF 9.0.0b03 (Note: MAPL only requires 8.6.1 at the moment)
- GFE 1.19.0
- gFTL 1.15.2
- gFTL-shared 1.10.0
- pFUnit 4.11.1 (optional)
- fArgParse 1.9.0 (if
-DBUILD_WITH_FARGPARSE=YES, default=YES) - pFlogger 1.16.1 (if
-DBUILD_WITH_PFLOGGER=YES, default=YES)
- UDUNITS2 2.28.8
Also, if you build with Baselibs and/or -DUSE_F2PY=ON, you should use (at least):
- ESMA_cmake v3.65.0
We recommend most external users set -DUSE_F2PY=OFF
We also require CMake 3.24 or higher.
From CHANGELOG.md
Fixed
- Fixed bug when one of the two names in an ExtData2G rule for a vector is contained is a substring in another item, for example UA;VA and EVAP
What's Changed
- Fixes #4158 by @bena-nasa in #4160
Full Changelog: v2.62.1...v2.62.2
Contributors
Assets 3
MAPL 2.62.1
This is a patch release of MAPL 2.62. This fixes a bug when regridding with grid that has PET with no DEs.
Testing shows this is zero-diff to 2.62.0
This version of MAPL was tested with:
- Intel Fortran Classic (
ifort) 2021.6.0 and 2021.13 - Intel Fortran (
ifx) 2025.2- NOTE:
ifx2025.2 support is currently a bit of a work-in-progress as it requires changes in upstream GFE libraries (pFlogger and pFUnit) as well as ESMA_cmake v3.65.0. Once releases are made of the GFE libraries, this will become simpler.
- NOTE:
- GCC 13.2.0, 14.2.0, and 15.2.0
- NAG 7.2
The libraries this version of MAPL is currently tested with are below.
- Baselibs 8.19.0
- netcdf-c 4.9.2
- netcdf-fortran 4.6.1
- ESMF 9.0.0b03 (Note: MAPL only requires 8.6.1 at the moment)
- GFE 1.19.0
- gFTL 1.15.2
- gFTL-shared 1.10.0
- pFUnit 4.11.1 (optional)
- fArgParse 1.9.0 (if
-DBUILD_WITH_FARGPARSE=YES, default=YES) - pFlogger 1.16.1 (if
-DBUILD_WITH_PFLOGGER=YES, default=YES)
- UDUNITS2 2.28.8
Also, if you build with Baselibs and/or -DUSE_F2PY=ON, you should use (at least):
- ESMA_cmake v3.65.0
We recommend most external users set -DUSE_F2PY=OFF
We also require CMake 3.24 or higher.
From CHANGELOG.md
Fixed
- Fixed bug when regridding with grid that has PET with no DEs
What's Changed
- v2: Update INSTALL.md by @mathomp4 in #4093
- Auto GitFlow - main → develop by @github-actions[bot] in #4095
- v2: Add NAG CI Runner by @mathomp4 in #4106
- fixes #4118 by @bena-nasa in #4119
- Prepare for 2.62.1 release by @mathomp4 in #4122
Full Changelog: v2.62.0...v2.62.1
Contributors
Assets 3
MAPL 2.62.0
This is a minor release of MAPL 2 which adds support for ifx 2025.2.
However, this support is not "instantaneous" at the moment. To build MAPL with ifx 2025.2, users will need to use ESMA_cmake v3.65.0 (which is what components.yaml points to) as well as updates in pFlogger (and its dependency yaFyaml) and pFUnit. These changes are all to workaround a preprocessor bug in ifx 2025.2 that will be fixed in ifx 2025.3. Releases of the GFE libraries will be forthcoming once full testing is done.
Note: @climbfuji has been able to update spack to make builds with ifx 2025.2 under spack easy until the GFE is updated.
All tests show this is zero-diff to MAPL 2.61
This version of MAPL was tested with:
- Intel Fortran Classic (
ifort) 2021.6.0 and 2021.13 - Intel Fortran (
ifx) 2025.2- NOTE:
ifx2025.2 support is currently a bit of a work-in-progress as it requires changes in upstream GFE libraries (pFlogger and pFUnit) as well as ESMA_cmake v3.65.0. Once releases are made of the GFE libraries, this will become simpler.
- NOTE:
- GCC 13.2.0, 14.2.0, and 15.2.0
- NAG 7.2
The libraries this version of MAPL is currently tested with are below.
- Baselibs 8.19.0
- netcdf-c 4.9.2
- netcdf-fortran 4.6.1
- ESMF 9.0.0b03 (Note: MAPL only requires 8.6.1 at the moment)
- GFE 1.19.0
- gFTL 1.15.2
- gFTL-shared 1.10.0
- pFUnit 4.11.1 (optional)
- fArgParse 1.9.0 (if
-DBUILD_WITH_FARGPARSE=YES, default=YES) - pFlogger 1.16.1 (if
-DBUILD_WITH_PFLOGGER=YES, default=YES)
- UDUNITS2 2.28.8
Also, if you build with Baselibs and/or -DUSE_F2PY=ON, you should use (at least):
- ESMA_cmake v3.65.0
We recommend most external users set -DUSE_F2PY=OFF
We also require CMake 3.24 or higher.
From CHANGELOG.md
Fixed
- CMake workaround for ifx 2025.2
- NOTE: Requires ESMA_cmake v3.65.0 for ifx 2025.2 support as well as updates in GFE not yet in Baselibs
Changed
- Update
components.yamlESMA_envv5.14.0- Update to Baselibs 8.19.0
- esmf 9.0.0b03
- curl 8.16.0
- Update to Baselibs 8.19.0
ESMA_cmakev3.65.0- Workaround for ifx 2025.2
What's Changed
- v2: Try and repeat failed ctest by @mathomp4 in #4077
- v2: Fixes for ifx 2025.2 support by @mathomp4 in #4084
- v2: Prepare for 2.62 release by @mathomp4 in #4087
- GitFlow: Merge Develop into Main by @mathomp4 in #4089
Full Changelog: v2.61.0...v2.62.0
Contributors
Assets 3
MAPL 2.59.1
This is a patch on MAPL 2.59 to allow for use with ESMF v9. ESMF v9 is not required for this release, but rather changes in ESMF v9.0.0b03 are needed in MAPL3 development.
The release is necessary due to an internal change in ESMF v9 that exposed an issue with the deprecated ESMF_Attribute still used in MAPL2 (see esmf-org/esmf#493). As such we now detect at runtime if ESMF v9 is being used and use the workaround code.
All tests with GEOSgcm show this to be zero-diff.
This version of MAPL was tested with:
- Intel Fortran Classic (
ifort) 2021.6.0 and 2021.13 - Intel Fortran (
ifx) 2025.0 - GCC 13.2.0 and 14.2.0
- NAG 7.2
The libraries this version of MAPL is currently tested with are below.
- Baselibs 7.33.0
- netcdf-c 4.9.2
- netcdf-fortran 4.6.1
- ESMF 8.8.1 (Note: MAPL only requires 8.6.1 at the moment)
- GFE 1.16.0
- gFTL 1.15.2
- gFTL-shared 1.10.0
- pFUnit 4.11.1 (optional)
- fArgParse 1.9.0 (if
-DBUILD_WITH_FARGPARSE=YES, default=YES) - pFlogger 1.16.1 (if
-DBUILD_WITH_PFLOGGER=YES, default=YES)
- UDUNITS2 2.28.8
Also, if you build with Baselibs and/or -DUSE_F2PY=ON, you should use (at least):
- ESMA_cmake v3.62.1
We recommend most external users set -DUSE_F2PY=OFF
We also require CMake 3.24 or higher.
From CHANGELOG.md
Changed
- Added support for ESMF 9
- Requires
#ifdefto support changes in deprecatedESMF_AttributeAPI in ESMF 9
- Requires
- Improved some error statements
Full Changelog: v2.59.0...v2.59.1
MAPL 2.61.0
This minor release of MAPL adds support for ESMF v9. ESMF v9 is not required for this release, but rather changes in ESMF v9.0.0b03 are needed in MAPL3 development.
The release is necessary due to an internal change in ESMF v9 that exposed an issue with the deprecated ESMF_Attribute still used in MAPL2 (see esmf-org/esmf#493). As such we now detect at runtime if ESMF v9 is being used and use the workaround code.
All tests with GEOSgcm show this to be zero-diff.
This version of MAPL was tested with:
- Intel Fortran Classic (
ifort) 2021.6.0 and 2021.13 - Intel Fortran (
ifx) 2025.0 - GCC 13.2.0 and 14.2.0
- NAG 7.2
The libraries this version of MAPL is currently tested with are below.
- Baselibs 8.18.0
- netcdf-c 4.9.2
- netcdf-fortran 4.6.1
- ESMF 8.9.0 (Note: MAPL only requires 8.6.1 at the moment)
- GFE 1.19.0
- gFTL 1.15.2
- gFTL-shared 1.10.0
- pFUnit 4.11.1 (optional)
- fArgParse 1.9.0 (if
-DBUILD_WITH_FARGPARSE=YES, default=YES) - pFlogger 1.16.1 (if
-DBUILD_WITH_PFLOGGER=YES, default=YES)
- UDUNITS2 2.28.8
Also, if you build with Baselibs and/or -DUSE_F2PY=ON, you should use (at least):
- ESMA_cmake v3.64.0
We recommend most external users set -DUSE_F2PY=OFF
We also require CMake 3.24 or higher.
From CHANGELOG.md
Changed
- Added support for ESMF 9
- Requires
#ifdefto support changes in deprecatedESMF_AttributeAPI in ESMF 9
- Requires
- Improved some error statements
What's Changed
- A few error message improvements by @mmanyin in #4056
- v2: ESMF 9 Support by @mathomp4 in #4062
- v2: Prepare for 2.61 release by @mathomp4 in #4065
- GitFlow: Merge develop into main for 2.61 release by @mathomp4 in #4068
Full Changelog: v2.60.0...v2.61.0
Contributors
Assets 3
MAPL 2.60.0
This release of MAPL has fixes and updates. Among the fixes are ones for NRL solar constant reads (when cycling over Cycle24) and fixes for NVHPC builds.
There are also two major changes. First, the default VERSION for HISTORY.rc is now 1 instead of 0. Nearly all GEOS History RC files have had VERSION: 1 at the top for many years and rely on functionality from it (e.g., using GRID_LABELS to specify grids). However, it was found that when this was accidentally left off by a user, the resulting run behavior was confusing as the symptoms were not obvious.
The other change is an enhancement of the MAPL Automatic Code Generator (ACG) which also fixes a bug in the Python. However, this bugfix can have an effect if a gridded component had a StateSpecs file that could trigger the bug. The issue was one of the "columns" a StateSpecs file can have is for how to handle fields restarts. MAPL has options like MAPL_RestartOptional and MAPL_RestartSkip that tell MAPL if a field is optional (if not present, that's okay) or skippable (if not present, okay, and do not write).
The ACG has allowed two ways for users to specify these options. You could use OPT or MAPL_RestartOptional and SKIP or MAPL_RestartSkip. The bug was that if a user didn't use the abbreviated option, the ACG did not emit the corresponding line.
An exemplar StateSpecs would be:
category: INTERNAL
#----------------------------------------------------------------------------------------
# VARIABLE | DIMENSION | Additional Metadata
#----------------------------------------------------------------------------------------
NAME | UNITS | DIMS | VLOC | UNGRIDDED | RESTART | ADD2EXPORT | FRIENDLYTO | LONG NAME
#----------------------------------------------------------------------------------------
SS |kg kg-1| xyz | C | self%nbins | MAPL_RestartOptional | T | DYNAMICS:TURBULENCE:MOIST | Sea Salt Mixing Ratio (bin %d)
DEEP_LAKES_MASK | 1 | xy | N | | MAPL_RestartSkip | F | | Deep Lakes Mask
Here we have two fields one, SS is MAPL_RestartOptional, and DEEP_LAKES_MASK is MAPL_RestartSkip. The difference in emitted code between MAPL 2.59 (and earlier) and MAPL 2.60 is:
@@ -15,6 +15,7 @@
& dims=MAPL_DimsHorzVert, &
& vlocation=MAPL_VlocationCenter, &
& ungridded_dims=[self%nbins], &
+ & restart=MAPL_RestartOptional, &
& add2export=.true., &
& friendlyto='DYNAMICS:TURBULENCE:MOIST', &
& long_name='Sea Salt Mixing Ratio (bin %d)', &
@@ -26,6 +27,7 @@
& units='1', &
& dims=MAPL_DimsHorzOnly, &
& vlocation=MAPL_VlocationNone, &
+ & restart=MAPL_RestartSkip, &
& add2export=.false., &
& long_name='Deep Lakes Mask', &
& _RC)MAPL 2.60 now correctly emits the restart= line. Now, for fields that use MAPL_RestartOptional instead of OPT the bug was a no-op. In MAPL, if there is no restart= line, the default to do is MAPL_RestartOptional. But the other MAPL_Restart* options would be affected.
For GEOSgcm, the only StateSpecs file that is affected is SS2G_StateSpecs.rc. So for GEOSgcm, when you compare a run with MAPL 2.59 to MAPL 2.60, the ss_internal_checkpoint will not have the DEEP_LAKES_MASK variable since it is now correctly skipped.
NOTE: The presence of DEEP_LAKES_MASK does not affect the state of the model. This mask is recomputed on each run and is constant.
This version of MAPL was tested with:
- Intel Fortran Classic (
ifort) 2021.6.0 and 2021.13 - Intel Fortran (
ifx) 2025.0 - GCC 13.2.0 and 14.2.0
- NAG 7.2
The libraries this version of MAPL is currently tested with are below.
- Baselibs 8.18.0
- netcdf-c 4.9.2
- netcdf-fortran 4.6.1
- ESMF 8.9.0 (Note: MAPL only requires 8.6.1 at the moment)
- GFE 1.19.0
- gFTL 1.15.2
- gFTL-shared 1.10.0
- pFUnit 4.11.1 (optional)
- fArgParse 1.9.0 (if
-DBUILD_WITH_FARGPARSE=YES, default=YES) - pFlogger 1.16.1 (if
-DBUILD_WITH_PFLOGGER=YES, default=YES)
- UDUNITS2 2.28.8
Also, if you build with Baselibs and/or -DUSE_F2PY=ON, you should use (at least):
- ESMA_cmake v3.64.0
We recommend most external users set -DUSE_F2PY=OFF
We also require CMake 3.24 or higher.
From CHANGELOG.md
Fixed
- Fix NRL Solar Constant read routine for cycle-Cycle24 option
- Change a few keyword names in sampler for consistency with HISTORY
- Fixes for NVHPC: Move some subroutines in
MAPL_MaskModfrom submodule to module - Fix for NAG + macOS Arm which does not support IEEE halting properly
Changed
- Updated the default
VERSIONin History to1(which has been the effective default inHISTORY.rcfor some time) - Update
components.yamlESMA_envv5.13.0- Update to Baselibs 8.18.0
- ESMF 8.9.0
- curl 8.15.0
- NCO 5.3.4
- CDO 2.5.3
- nccmp 1.10.0.0
- Removed szip, added libaec
- Note: To use libaec correctly, users should use
ESMA_cmakev3.63.0/v4.20.0 or higher
- Note: To use libaec correctly, users should use
- Update to Baselibs 8.18.0
ESMA_cmakev3.64.0- Support for libaec
- Add
FindISSM.cmake
- Update CI to use Baselibs 8.18.0
- Allow row lookups to return key if key in values (ACG2)
What's Changed
- Bump actions/checkout from 4 to 5 by @dependabot[bot] in #3977
- Sampler code: change a few variable names by @metdyn in #3985
- Bump devops-infra/action-pull-request from 0.6.0 to 0.6.1 by @dependabot[bot] in #3986
- v2: Fix for NAG halting on macOS Arm by @mathomp4 in #4007
- v2: Update to ESMA_env 5.13.0, ESMA_cmake 3.64.0 by @mathomp4 in #4028
- Bump actions/setup-python from 5 to 6 by @dependabot[bot] in #4035
- Bump actions/stale from 9 to 10 by @dependabot[bot] in #4036
- v2: Update Spack CI by @mathomp4 in #4039
- Allow lookups to return as value by @darianboggs in #4040
- Closes #4024. Change History to default VERSION: 1 by @mathomp4 in #4025
- Fixes #2623. Use modulo math for off-table Cycle24 recycling by @mathomp4 in #3972
- v2: Fixes for NVHPC Build by @mathomp4 in #3989
- v2: Prepare for 2.60 release by @mathomp4 in #4048
- GitFlow: Merge develop into main for 2.60 release by @mathomp4 in #4051
Full Changelog: v2.59.0...v2.60.0