Conversation
|
As a nomenclature thing, I think I prefer calling this mirroring instead. |
Had the same thought when opening the PR 🙂 Will refactor it. |
Co-authored-by: Natalie Klestrup Röijezon <nat@nullable.se>
Co-authored-by: Natalie Klestrup Röijezon <nat@nullable.se>
Co-authored-by: Natalie Klestrup Röijezon <nat@nullable.se>
Co-authored-by: Natalie Klestrup Röijezon <nat@nullable.se>
Co-authored-by: Natalie Klestrup Röijezon <nat@nullable.se>
Co-authored-by: Natalie Klestrup Röijezon <nat@nullable.se>
2bf81bd to
f85102d
Compare
f85102d to
694e5c7
Compare
Co-authored-by: Natalie Klestrup Röijezon <nat@nullable.se>
|
@nightkr I would now go ahead and update all the |
|
Sounds good to me. |
…into feat/patchable-init-forking
|
@nightkr This is done now. Product-level configs are all there and I removed |
|
All versions seem to check out fine ( |
|
@nightkr I noticed on more thing 🙁 Some builds would fail, because we use |
|
Ahh, right. Good catch. |
|
Hm, that would mean losing caching for all versions when any version is modified, though. It should be enough to copy |
…roduct config" This reverts commit 6148608.
|
@nightkr Reverted and fixed by copying only the product config |
Fixes #1049
This adds the
--mirroredflag topatchable init. It assumes that a mirror repository already exist, pushes the ref specified via--baseto the mirror and configures the mirrorer repo asupstreamin patchable.toml.As discussed with @lfrancke, we want to fork the repos in the
stackabletechorg and use forks instead of clones (the latter decision was not a strong opinion, I decided to go with forks since they would facilitate back-contributions).I created mirror repositories (forks in our case) for all our products and synced the tags for all versions we use. I updated all patchable.toml files to use the mirror repositories (with the exception of Omid
1.1.3-SNAPSHOT, which is a temporary version we don't ship).I had to drop the usage of
FETCH_HEAD, since it caused problems when runningpatchable initfor a tag/branch that for some reason was already present in the local Git repo. The problem is thatFETCH_HEADis empty in that case, I think that's due to no changes being fetched.