[KimDaBa] Re: Search bar & CategoryEditor & Quick tagging (following)

[Jean-Michel FAYARD]

Me again, sorry for the spam ;-) 
~= ten days ago, I wrote this :


<===== ACT I =====>
* Has someone noticed the nice idea that William Holland made
while we were smiling at a joke from Jesper ?

> Secondly, as I look through the images I would like to be able to tag them in
> an 'ultra quick way' eg typing 't' or something, so that I can select the
> tagged ones later, and then perform some operation (setting key words, or in
> my case emailing them)

I didn´t understand at first, but that´s a rich idea : that would be
the viewer/slideshow version of "Images -> Limit view to selected".

Say I am watching my 300 pictures of my trip in the USA, and while
that slideshow I see that :
- this (picture #4 in the slideshow) picture is junk, delete it
- this one (#6) is wonderful, I need print them
- those are (#8, #11, #15) wonderful too
- Oh, here (#120) and there (#125) I forget someone on the picture. I
have to correct that
- ... whatever else

In Kimdaba you can stop the slideshow and change that and then restart
the slideshow each time. But William´s suggestion is far better, you
just select
'Right Click -> Mark picture as selected' or, more quickly, just type ´M´ and
let the slideshow continue.
When the slideshow ends or when you quit the viewer, the thumbnails
view shows only #4, #6, #8, #11, #15, #20 and #25 and you can do your
changes from there, all at one time.
</=========>

Probably a bad idea since nobody seemed to like it
Then I saw this entry in the todo list :


<===== ACT II =====>
- Add a way to mark images using a single key stroke, something like:
  select a bunch of images, start the viewer in the viewer SPACE set some
  options for the image, and go to the next image, while RET just goes to the
  next image. (That way you can easily go through hundreds of images putting
  the keyword "print", and later search for keyword print for sending images
  to printer)
  This could of course be extended so you could set keywords with
  different key bindings (which should be configurable):
  (c) - image is from Copenhagen
  (l) - image is from Las Vegas.


Sounds nice, huh ? But on the next line, Jesper noted :

         I doubt that extension will be very useful tho.

He´s right, because when you fire your viewer and start a slideshow,
two minutes after that, you need those handy "c" and "i" shortcuts,
one of those two things will happens :

  - you don´t have yet defined those shortcuts. So you want to do
    "Settings -> Configure Kimdaba", but your girlfriend says
    "Let´s just see the pictures for now, you will do that after"
    with a so lovely smile that you can not contradict her

  - you have already define the shortcut, but one month ago.
     Too bad, you already forget it, or didn´t remember that you
     changed this shortcut to means another property and do
     a mistake ;-)

So Jesper is right, that is a second bad idea

</=========>


<===== ACT III ====>
Now, let´s try to combine those two bad ideas to make a good one ;-)
You start the slideshow, you press like before (example from Jesper)
  (c) - image is from 'C'openhagen
  (l) - image is from 'L'as vegas.
but you can also press in my example :
  (p) - there is a 'P'erson missing on this one
  (j)  - this picture is 'J'unk, delete it


The point is that (like in my idea) 

1) YOU DID NOT CONFIGURE THAT BEFORE (and don´t have to),
  those shortcut are transparent for Kimdaba, they are just keystroke that
  are not binded yet with another action (everything except I D S Q
for now), and
  they mean whatever I want them to mean at this moment.
2) Those shortcuts don´t actually do anything like setting property Copenhagen
   for the image. Instead they sort out your pictures by setting a    
   property    "Key C" for a new temporary category "Shortcuts pressed".
   That is, after your slideshow, you are in a browser looking like this :

     |---------------------|---------------|
     | Shortcuts pressed   |  Count        |
     |---------------------|---------------|
     | Key C               |  10   images  |
     | Key L               |  4    images  |
     | Key P               |  3    images  |
     | Key J               |  25   images  |
     |---------------------|---------------| 

Since the shortcut does nothing, it´s not bad if you press a key by mistake,
just go back in the browser and that´s it.
But if you take our example, you just have to do now :

 	Click on "Key C"
	Ctrl-a, Ctrl-2, Type Copenhagen, Enter
	Go back in the browser
	Click on "Key P" (means 'Persons' in your head)
	Ctrl-a, You have every pictures where a person was missing, 
              Ctrl-1, type the names, Enter
	Go back in the browser
	Ctrl-a, Deselect the second one (not that bad after all !), 
              Edit->Delete selected  

Did I lost everybody, or does that sounds useful ?
</=========>

PS: I actually tried to implement this, but I´m just not good enough
at Qt/KDE programming yet :(