Closed mosabua closed 9 years ago
Log and continue already works but its has ugly stack traces... need to clean this up.
This is cleaned up now in terms of the log. But we are still doing duplicate deployments (overwrites) or we fail after trying. It would be faster to query target repo .. but that will depend on what runs there.. so is more complex than the brute force...
According to Igor we can still use the Aether ArtifactResolver ... just specify only the target repo. You could also ask for the metadata for all available vrsions .. but that only works if the remote is trustworthy in the sense that it provides the right info. Also that might not matter in our case. It might however still resolve the artifact (as in fully download it... and that would be bad..).
So maybe we still have to assemble the remote path and do a plain http request for it and abandon retrieval once download starts to avoid the time cost of a download..
Might not even be worth the time, especially if the files that are deployed are not big. They are uploaded first and then it fails... so for bigger files it should probably be done.
In progress..
Pushed to master
or for starters just log that deployment failed and continue