Jesper Pedersen (3a130ed4) at 30 Jul 18:36
Support annotating images from the viewer
Jesper Pedersen (a10902e5) at 30 Jul 15:03
Moved documentation to the docbook files.
Jesper Pedersen (d92f04a1) at 29 Jul 17:47
Add asignments to dialog for assigning keys
Jesper Pedersen (05905440) at 29 Jul 08:39
Fixed typo
I'm a bit afraid it will take focus away from the left side, which will be much smaller on the screen.
Jesper Pedersen (2d72cf41) at 29 Jul 08:34
Now it is possible to tokenize, annotate or (starting) nothing
Jesper Pedersen (4d339aa5) at 27 Jul 07:55
Fixes for the review
For the first part, I've added a help button to the key assigning dialog - after all help should only be used very seldom, so to include it all there would be a lot of visual noise.
I'll now try and merge the tokens and new assignment workflows in a separate commit.
They are somewhat different already, as the dialog one updates the list selects, while this updates and image info. I prefer not try to make them share code, as I'm pretty sure it will make it harder to understand.
I'll include this for tokens indeed.
commented above.
indeed, added.
Wrong line you referred to, right, should have been line 125. Removed completer there, but it is for sure needed on line 121
darn, thx
I actually thought about that for a long time. I was tagging a lot of photo while being on vacation with my laptop, and using that dialog (since I had forgot my time machine at home, so I couldn't get to the - at the time not yet invented - new way :-)
On my laptop it was super hard to get to shift-insert, as the insert key was hidden behind an FN key.
I would therefore propose we go for Ctrl+c in both places. However, having tried that, due to the use of line edits in the dialog, that will be impossible and un-intuiative too.
I expect that once people learn about this new workflow, they will only use the old dialog for mass adding tags, rather than for going though them one at a time.
So unless you feel strong about it, I'd like to keep it at Ctrl+c
The idea with the wall of text, is to make this feature more discoverable. Users see the help entry in the context menu and try it out. Once you are used to the feature you will just ignore it.
I've added a blank column to make it even more obvious there are two bindings per row. I've opted for two per row to make it easier to get an overview without the need for scrolling.
If you still can't live with it, then please suggest how to make it look instead.
Yup
The alternative would be that you either should confirm for each key binding, which would be annoying and slow this down. Another alternative would be that it should be saved together, but then there is a chance you forget to save it all together, which would be annoying too.
The OSD helps you realize what is on each key, and I personally found that my set of keys are relatively static - the members of my family, and for older pictures my two now dead dogs, and for newer my new dog.
Then in addition to that there are burst of "visitors" who I recognize fast and assign to keys temporarily. Worst case, if I forget who/what is on a key, then I simply press it, and if it wasn't what I expected, then I'll press it again to unassign.
Indeed, fixed.
Jesper Pedersen (3e71730f) at 25 Jul 07:11
Remove the need for a private header