Review available revisions for a specific object on restore; preserve user input

[saspus]

A very important feature request regarding restore experience:

Both, currently released UI and Beta Web UI in the Restore workflow force me to first select revision, and then browse to the file or folder I want to restore. This is not how human brain works. Virtually any time I want to restore something I definitely know what object I want restore but I’m often not sure which version of it do I need.

Therefore it is way more natural to first select an object and then select revision.

Presently, once I select revision and navigate to a file, and then realize that this was not the revision I wanted to restore – I can chose different one; but my path to the object is reset and I have to browse to that same objet again. This is highly annoying. This violates a very important UX principle - never lose data user created and inpu.

What I expect as a minimum: once I browsed to a specific file at certain revision, but then changed selected revision in the combo box above - my location in the hierarchy shall be preserved, as much as possible. I should not be forced to browse to the same file again.

Optimally, I would like to be able to browse for a file or folder, and then list all available revisions of that file or folder, with ability to QuickView known types of documents, and then select which one I want to restore.

This is a very important use case which for some reason vast majority of backup tools ignore, with very rare exceptions.

EXAMPLES:

1. Bad behavior: Restore workflow in the current web-ui beta:
   

   

2. Good behavior (with the room for improvement in the flow, but functionality is there): Version Explorer in Synology Cloud Station Backup
   

   

[DUser]

I fully second that. The discontinued CrashPlan also had a very good behaviour.

[michel]

I fully agree with this request.
Do we know if this is in the roadmap or not ?

[350c183cd72f8a081539]

I’d love to see this in the webgui. You can do something similar with the command line version, by listing all the files and grepping through them, but GUI access would be much more usable.

[rainforest1155]

Another former Crashplan user, I’m really missing better restore features on Duplicacy. Seeing the versions for a given file without having to manually switch revisions and then having to locate the file all over again would be really useful to have.

[Droolio]

Actually, you just reminded me of another missing feature that should be part of a backup tool - finding / locating files. This is very much related to the other, recently discussed, issue here: https://forum.duplicacy.com/t/how-can-i-find-which-chunks-make-a-file-and-the-other-way-around/6348/7?u=droolio

Related, because of the way Duplicacy works in separating chunk data from metadata - it becomes rather difficult to ‘search’ for a given file or folder based on mere filename, because that file name or even where it’s located in a directory structure, may change many times between snapshot revisions.

Duplicacy doesn’t directly track file renames and moves - it doesn’t care - it just backs up new data and de-duplication just happens, but it isn’t able to track revision history. So any search feature would be crude and unreliable - without the ability to cross-reference chunk data with revision history metadata.