Is there way to access fileās metadata?
Yes. The Dropzone element exposes each fileās metadata.
Alternatively a backend action is available to retrieve a stored fileās metadata as well.
Is there a way to split this into two options?
1st option would be āEnable Folder Uploadā (behaves like yesterdays version)
2nd option would be āRetain Folder Structureā (behaves like todays version)
That way it can still be like the version from yesterday where you can drop in a folder and it just uploads everything into one folder instead of replicating the structure?
It should currently behave as you described.
I uploaded a folder from Windows named level 1, with a folder in that called level 2, with the pdf in that,
Then when I upload in the workflow I first create a thing, and add that in front of the upload path (so every file is in itās own subfolder):
But itās keeping the level 1/level 2/file10.pdf in the upload name so itās showing in Wasabi it created level 1/level 2 as folders:
What I would expect if the āretain folder structureā can be turned off is it just uploads āfile10.pdfā under the first folder ā1661890054443x390713387688919000ā
Iām assuming itās uploading all those folders because the ālevel1/level2/ā is in the name and Wasabi interprets that has āmake a folderā?
Is that how it handled it yesterday maybe I didnāt notice?
I canāt reproduce the issue on my system. I tried with level1/level2/file to make sure.
Do you have a UNIX system/different browser at hand to try out? I suspect a platform-related bug.
Okay tried it on Win10 Chrome, Win10 Edge ,Ubuntu 22 Firefox and all do the same thing.
Interesting but I can drag a folder right into the Dropzone and it uploads like I was showing earlier.
But If I click on the drop zone to browse files, on Windows and Ubuntu it wants me to browse to a specific file it doesnāt want to upload a folder
But If I select a folder + a file it will upload all of it with the structure problem stillā¦ I just tried that on Ubuntu.
Maybe this is intended behavior? Idk 
Would you mind to DM me a loom/screencast of the issue with the settings along with demo page, ideally with editor access?
Just to make sure. The āRetain folder structureā option must be unchecked to upload all files and folders without retaining their structure.
Yea that would make sense. Itās almost like it could be 3 separate options:
āUpload Multiple Filesā allows selecting multiple files
āEnable Folder Uploadā allows selection of folders
āRetain Folder Structureā would do what itās doing now but for my case I would uncheck it so it will search all folders but only upload the files with their names. (no level1/level2/)
In the latest version only those are implemented:
Oops sorry I was on 1.4.0 my editor didnāt show a new version until I refreshed just now. Iāll try the 1.4.2 right now.
Looks good thank you so much. Iāll compile my remaining ideas and check with you if they are possible. Right now I think whatās left is:
- Rename files (technically would this enable a move function as well since you could rename it with a new path?)
- Upload as a āpublicā file so it can have a direct URL
I know you said itās not possible but keep it on your suggestion list please, I just timed it, it took 10 seconds after dragging in a 1MB PDF for it to initialize the upload before it could start uploading. It was quick from that point on but when changing pages it always takes a long time. 
Also hereās another one:
Is it possible for this action to have a field to specify a path? So it could retrieve all the URLās inside a subfolder instead of bucket wide? In this case Iām looking at it in the backend.
Done.
Copy action has been implemented. Renaming is therefore available by the combination of Copy then Delete action.
Sweet Iāll try it out. Is there something in place so the copy starts, then the delete wonāt start until the copy is done? You know how Bubble doesnāt really do things one step at a timeā¦
If the Copy action does not return the file path / URL, then it failed. You may use it as a test for the next action.
Yep looks good 
On the āCopy fileā action it seems like itās missing the āTarget Bucket Regionā field but it allows a different Bucket name.
Also on upload I store the resultās UUID in my database, when copying/deleting is the new fileās UUID different? Is the UUID even needed when I store it?