Main changes from last release (3.2.8 -> 3.3.0) This release is primarily aimed at performance and resource usage improvements. We have optimized the memory usage in CMOR and PrePARE. In addition, we changed the file IO library for PrePARE from CDMS2 to CDUNIF to achieve an order of magnitude speedup in file validation operations. These changes should achieve a considerable speedup in ESGF data publication, and a considerable reduction in memory usage enabling better parallelisation and scaling of multiple concurrent write operations Accelerate file validation by PrePARE from 2s to 0.1s (IPSL request) Fix PrePARE/CMOR memory issue where table references were not dynamically assigned (MOHC request) remove fixed double array to dynamic double due to number of objects increasing Optimize PrePARE file validation speed to reduce validation operations Call Cdunif.so library directly instead of cdms2 "long_name" variable attribute is no longer being validated by PrePARE (CERFACS request) Update "license" REGEX to allow "/" at the end of URL (GFDL request) Update "Conventions" REGEX to allow "CMIP-6.0" up to the current "CMIP-6.2" release Update grid tables remove standard_name to vertices_latitude and vertices_longitude for CF-1 and CMIP-6.2 compliance CMOR 3.3.0 documentation http://cmor.llnl.gov or in pdf form at http://cmor.llnl.gov/pdf/ DOI (source code) https://doi.org/10.5281/zenodo.1115862 Github (source code) https://github.com/PCMDI/cmor/releases/tag/CMOR-3.3.0 Conda installation To install cmor into your root anaconda environment conda install -c conda-forge -c pcmdi cmor Or to generate a dedicated anaconda environment conda create -n cmor3.3.0 -c conda-forge -c pcmdi cmor Using this cmor meta yaml file conda env create -n cmor3.3.0 -f cmor-3.3.0.yml