How do I change the remote of gin sibling?

[camiscaff]

Current remote is [[email protected]:/LAAC-LSCP/tsimane2018.git (git)]
we want to be : [[email protected]:/LAAC-LSCP/tsimanec2018.git (git)] #notice the addition of the "c"

[camiscaff]

Another question: (ChildProjectVenv) (base) cscaff@MN058AD42G86 tsimanem2018 % datalad push --to gin
[ERROR ] Unknown push target 'gin'. Known targets: 'origin'. [publish(/Users/cscaff/tsimanem2018)]
publish(error): . [Unknown push target 'gin'. Known targets: 'origin'.]
(ChildProjectVenv) (base) cscaff@MN058AD42G86 tsimanem2018 % datalad push
publish(ok): . (dataset) [refs/heads/master->origin:refs/heads/master [new branch]]
(ChildProjectVenv) (base) cscaff@MN058AD42G86 tsimanem2018 % datalad siblings
.: here(?) [git]
.: origin(?) [[email protected]:/LAAC-LSCP/tsimanem2018.git (git)]

How can I harmonize the behavior of the two folders tsimanem2018 and tsimanec2018 so I can push identically ( with "push --to gin" or simply "push")

[lucasgautheron]

To configure the "origin" sibling to point towards <newurl>:

datalad siblings configure -s origin --url <newurl>

If the sibling does not exist, the command will create it.

By default, when cloning/installing a dataset from a source, the sibling is automatically named origin and it is configured as the default sibling. Therefore I would recommend always configuring the main sibling with the name "origin" always rather than "gin" locally.

More generally, I strongly advise using these procedures to create new LAAC or EL1000 datasets: https://github.com/LAAC-LSCP/datalad-procedures

These procedures automatically configure the siblings and the structure of the dataset.

More datalad documentation here: http://docs.datalad.org/en/stable/generated/man/datalad-siblings.html

[alecristia]

thank you! We had indeed forgotten about procedures. So now the situation is that we have a local version that was not created using procedures, and which has been pushed to GIN. There is not a lot of work on it, so we can re-start from scratch.

Would you suggest that? And if so, would it really be from scratch: force remove local copy, delete copy on GIN, and start over?

Thanks in advance!

[lucasgautheron]

I would recommend that. It's usually easier to work from a clean state, unless you feel very confident you can tweak your way back to that state!

I would suggest to:

1. Remove the datasets from GIN (delete them from GIN's interface)
2. Keep a copy of your files
3. Recreate a dataset with these procedures
4. Fill these new datasets with the backup of your past progress

[alecristia]

5. Have a pint

[lucasgautheron]

Yes (preferably in that order)