Description
JabRef 5.2--2020-12-24--6a2a512
Windows 10 10.0 amd64
Java 14.0.2
- Mandatory: I have tested the latest development version from http://builds.jabref.org/master/ and the problem persists
If the .bib
file is shared on Dropbox/OneDrive/WebDAV with a somewhat slower connection (i.e. slower than an SSD writing speed), the auto-backup feature reverts back to the backup version, losing the newly 'saved' file in the updated .bib
file. I am using a stable version 5.2.
Steps to reproduce the behavior:
- Open a bib file from Dropbox/OneDrive with 1000 entries
- Create a new group, Add a new Entry and Save it
- Close the file/tab
- Observe the creation of
filename.bib.bak
,filename.bib.sav
, andfilename.bib.sav.tmp
on the filesystem - Do not close JabRef, as within a minute the bib file will reopen (!!!) without any user interaction
- The problem: the file is not the newly saved file with including the new entry, but the autosaved version without the new entry. I.e. if you save this file on closing, you are losing the new entry from the previous save without any warning.
- If you reject this save, the tab with the file does not open anymore.
This is IMHO a serious bug, as one can lose entries. I checked the online documentation how things intend to work, however, this is somehow not ideal on shared drives I guess. Could you please investigate this further?
Besides the obvious bug or automatic reopening of the auto-backup file, I was wondering whether with large datasets (with 100s of entries) it would be advisable to inform the user on saving of some basic stats, i.e. about the change in total amount of entries, how many new entries are going to be saved, and how many entries were edited, eventually deleted. This should be doable as a result of comparing the auto-backup files with the newly saved one. With this approach it can be ensure that the user is fully informed and want to commit the changes.
Metadata
Metadata
Assignees
Labels
Type
Projects
Status