Exports from record buckets are cached problematically by Chrome
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Evergreen |
Confirmed
|
Medium
|
Unassigned |
Bug Description
EG 3.1
When you export records via record buckets in the web client, the export is cached. If you add different records to the same bucket and try to export the new records using the same format/encoding, it will generate new export files, but each file contains the records from your initial export. Effectively, all subsequent exports contain the wrong records.
1. Retrieve a record (record #1) in the client and add it to a bucket via the Actions menu.
2. Go to Cataloging > Record Buckets and view the bucket. It shows record #1.
3. Select the record and export via Actions menu. The download file contains record #1.
4. Remove record #1 from the bucket.
5. Add a new record (record #2) to the same bucket.
6. View the bucket again. It shows record #2.
7. Select record #2 and export it using the same format and encoding as before. This generates a new download file, but the file contains record #1.
Presumably the browser is caching the download file/URL and not bothering to re-download anything until the cache expires. Adding a cache-busting key to the filename (either for each download attempt or when the export contents have changed) would probably avoid the problem.
I'm only seeing this in Chrome (version 73.0.3683.86), not in the equivalent version of Chromium. I haven't tried Firefox.
tags: | added: buckets webstaffclient |
tags: |
added: buckets-record removed: buckets |
Confirmed that this still happens in 3.6