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
Vesper currently does not allow more than one clip with the same detector, recording channel, and start time in an archive. This is accomplished via a multicolumn uniqueness constraint on the clip table. I included the constraint to prevent duplicate clips from being created by accidentally running a particular detector on a particular recording more than once. The problem with that is that there are legitimate reasons for creating more than one clip for a given processor, recording channel, and start time. As one example, an upcoming detector sometimes creates multiple clips with the same start time for overlapping detections at multiple taxonomic levels. So I think it would be best to remove the constraint and rely on another means of dealing with undesired duplicate clips. Right now it's possible to delete a job from the database, which also deletes the job's clips, but that is not a good solution since we'd like to maintain a complete record of the jobs that have been run. A better solution might be a command that deletes clips created by a particular job, perhaps only for certain sensors. It would also be desirable to add a command that can detect duplicate clips in a database, regardless of which job created them, and optionally delete them.
The text was updated successfully, but these errors were encountered:
Vesper currently does not allow more than one clip with the same detector, recording channel, and start time in an archive. This is accomplished via a multicolumn uniqueness constraint on the clip table. I included the constraint to prevent duplicate clips from being created by accidentally running a particular detector on a particular recording more than once. The problem with that is that there are legitimate reasons for creating more than one clip for a given processor, recording channel, and start time. As one example, an upcoming detector sometimes creates multiple clips with the same start time for overlapping detections at multiple taxonomic levels. So I think it would be best to remove the constraint and rely on another means of dealing with undesired duplicate clips. Right now it's possible to delete a job from the database, which also deletes the job's clips, but that is not a good solution since we'd like to maintain a complete record of the jobs that have been run. A better solution might be a command that deletes clips created by a particular job, perhaps only for certain sensors. It would also be desirable to add a command that can detect duplicate clips in a database, regardless of which job created them, and optionally delete them.
The text was updated successfully, but these errors were encountered: