You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
@mathomp4 For the CUPID automated testing it may be best if FindBaselibs has override options that point to specific places for each of the (non-GEOS) dependencies: FLAP, pFUnit, gFTL, gFTL-shared, Netcdf ... This would also be useful for testing updates on individual dependencies without requiring you to roll out a new baselib build each time.
The idea would be that if a user has specified say -DESMF_DIR=, then the find logic for ESMF will default to there instead of BASEDIR.
The alternative is to have MAPL fully automate the cloning and installation of all dependencies in standalone mode. This is probably somewhat good from the PoV of @rsdunlapiv, but rather expensive in practice if you already have a build of the various dependencies.
And of course a hybrid approach could be done where MAPL automates the build of the dependencies that are missing, but ... that would require significant changes or would completely sidestep the logic in ESMA_Cmake.
The text was updated successfully, but these errors were encountered:
@mathomp4@LiamBindle I actually think that this ticket was already completed, but simply did not automatically close. We certainly demonstrated something to the ESMF team using a narrow ESMF. Liam's fix might be more general.
@mathomp4 For the CUPID automated testing it may be best if FindBaselibs has override options that point to specific places for each of the (non-GEOS) dependencies: FLAP, pFUnit, gFTL, gFTL-shared, Netcdf ... This would also be useful for testing updates on individual dependencies without requiring you to roll out a new baselib build each time.
The idea would be that if a user has specified say -DESMF_DIR=, then the find logic for ESMF will default to there instead of BASEDIR.
The alternative is to have MAPL fully automate the cloning and installation of all dependencies in standalone mode. This is probably somewhat good from the PoV of @rsdunlapiv, but rather expensive in practice if you already have a build of the various dependencies.
And of course a hybrid approach could be done where MAPL automates the build of the dependencies that are missing, but ... that would require significant changes or would completely sidestep the logic in ESMA_Cmake.
The text was updated successfully, but these errors were encountered: