on Azure:
sudo apt install git tmux make python3-venv datalad
tmux
git clone https://github.com/mikapfl/unfccc_di_data
cd unfccc_di_data
rm -rf venv
make venv
source venv/bin/activate.fish (in case you use bash without the '.fish')
datalad update -s ginhemio --how merge
datalad rerun …

datalad siblings add -s ginhemio --url https://gin.hemio.de/CR/unfcc_di_data --pushurl git@gin.hemio.de/CR/unfcc_di_data
git config remote.ginhemio.annex-ignore false

datalad push --to ginhemio --data anything

At home:
datalad update -s ginhemio --how update
datalad update
git pull

if you want to compare differences, get the old and new CSVs, then run (in fish):
for i in old/*.csv; echo $i; csvdiff --ignore-columns 0 -p 1,2,3,4,5,7 -o word-diff $i  new/$(basename $i) > diff_$(basename $i); end
you need https://github.com/aswinkarthik/csvdiff for that.

if you are happy, start making new version on zenodo, get pre-reserved doi
put new doi, new info into datacite.yml and README.md
datalad save
datalad push
datalad push --to origin
datalad push --to ginhemio --data anything
datalad export-archive -t zip 'data-2022-10-12.zip'
zip parquet-only.zip data/annexI/*.parquet data/non-annexI/*.parquet
upload new data (including parquet-only.zip) to zenodo, publish
new version of unfccc_di_api