You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is your feature request related to a problem? Please describe.
The file UUID generation task occurs in three separate places: Assign file UUIDs to objects, Assign file UUIDs to submission documentation, and Assign file UUIDs to metadata. In Archivematica, with its two-stage processing, the file UUIDs are generated during transfer and the metadata and submission documentation UUIDs are generated during ingest. This makes sense for Archivematica because submission documentation and metadata files can be added during backlog and appraisal, between the two stages.
a3m no longer has two-stage processing, so there is no opportunity for users to add or change the submission documentation and metadata files. It seems unnecessary to separate out the jobs in this way, when they all accomplish the same basic task.
This is an extension of #702.
Describe the solution you'd like
Condense the file UUID tasks into one task (could just be called "Assign file UUIDs") that covers everything in the transfer (all digital objects as well as anything contained in a metadata or submission documentation folder - in a bag, this would be everything in the data directory).
This helps to reduce the overall number of jobs in a3m, and contributes to dismantling the two-stage workflow inherited from Archivematica.
Additional context
Note that there is also a UUID generation task for directories, but I'm happy to leave that as a separate task, since it's something that can be configured.
The text was updated successfully, but these errors were encountered:
I can write more issues, but this same logic could also be applied to:
filename change
checksum generation
characterize and extract metadata
file format identification (in this case, we'd have to apply a blanket rule, i.e. all metadata/submission documentation should be ID'd along with the payload)
There may be an opportunity to pull this out of a3m entirely and have it be something implemented in enduro. I think this might be a good starting point.
Is your feature request related to a problem? Please describe.
The file UUID generation task occurs in three separate places:
Assign file UUIDs to objects
,Assign file UUIDs to submission documentation
, andAssign file UUIDs to metadata
. In Archivematica, with its two-stage processing, the file UUIDs are generated during transfer and the metadata and submission documentation UUIDs are generated during ingest. This makes sense for Archivematica because submission documentation and metadata files can be added during backlog and appraisal, between the two stages.a3m no longer has two-stage processing, so there is no opportunity for users to add or change the submission documentation and metadata files. It seems unnecessary to separate out the jobs in this way, when they all accomplish the same basic task.
This is an extension of #702.
Describe the solution you'd like
Condense the file UUID tasks into one task (could just be called "Assign file UUIDs") that covers everything in the transfer (all digital objects as well as anything contained in a metadata or submission documentation folder - in a bag, this would be everything in the data directory).
This helps to reduce the overall number of jobs in a3m, and contributes to dismantling the two-stage workflow inherited from Archivematica.
Additional context
Note that there is also a UUID generation task for directories, but I'm happy to leave that as a separate task, since it's something that can be configured.
The text was updated successfully, but these errors were encountered: