[Meta] scanner handling
(This ticket is here to keep track of all scanner related issue needing ux improvement)
Description
Most scanners offer to scan only one page at a time (flatbed), while some allow to scan a bunch of pages at once (flatbed). A few even support scanning recto & verso. The GUI should provide the user with options relevant to the device in use.
Issues
- [] Two-sided scanning #535
- [] "Scan from ADF" (batch mode) should support source selection (duplex) #554
- [] Merge recto and verso pages #579
Just one comment : I don't think the part where the label list appear from the right is really relevant. In a perfect world, label should be automagically guessed correctly. Also, currently they appear on the left side .. :/
Just one comment : I don't think the part where the label list appear from the right is really relevant. In a perfect world, label should be automagically guessed correctly. Also, currently they appear on the left side .. :/
Thanks for this comment. Even in a perfect world, the user would have to create & edit tags before paperwork is able to automatically associate them with papers. So the question is: why moving the pane from the left to the right? I see 2 reasons to do so.
-
I am not aware of software other than Paperwork that would show the properties of a document on the left of it. Activating a "details" button to have file properties/metadata displayed in a right pane seems a more common pattern. Some examples here: Nextcloud, plus.google.com, Bear...
-
As we are adding new features (overview, export queue, saved searches...) we need to keep things simple. What we did in the latest wireframes experiments is to split the interface in a serie of vertical panes. The idea is to build a simple workflow, going from the general (left) to the specific (right). This is illustrated here and there.
scan from feeder
Why is it necessary to tell the application how many pages it has to scan? Doesn't the device report the end/last page - can I help by reading out some logs? With regard to automagically, paperwork would serve me one-page-after-the-other and either I 'accept' paperworks Label suggestion "a tick to click" or I click 'Edit'.
PS: 3-column layout with 16:10 monitor makes sense to me. Having properties on the right as well.
Actually, I'm considering removing the whole multi-scan dialog, and just assume that the user has put 1 document in the feeder:
- If "New document" is selected --> it creates a new one.
- If another document is selected --> it appends the pages.
It would make the behavior pretty much consistent with when we scan single pages.
1 document in the feeder
What was the reason for this dialog in the first place? Which problem did it solve?
If you throw three invoices on the ADF, 1 page, a 2 page'r and a 3 page'r. I play your scenario, but instead of 'another' I would call it 'additional': 1 page = click New Document 2 pages = click New Document (recto), additional (verso) 3 pages = click New Document (recto), additional (recto), additional (verso) now the scanner sends: end/last page What I am questioning/curious will it work without the dialog and how to skip a white page?
What was the reason for this dialog in the first place? Which problem did it solve?
It was mostly focused on first time users : They usually have a lot of documents to scan --> they take 10 documents, put it in the scanner, configure the mutli-scan dialog, launch the process, and go take a coffee. But frankly, it seems now like a poor idea. It would actually be much simpler to just scan pages until the feeder is empty, and then let the users rearrange the pages between the (new or not) documents.