Skip to main content

⚠️ Please note that this topic or post has been archived. The information contained here may no longer be accurate or up-to-date. ⚠️

Bug with keywords

Comments

6 comments

  • Permanently deleted user

    My workaround is to not use hierarchical keywords.  Well, I use them, but manually.   In the example above I'd have to manually select both Dogs and Animals.   I started doing that because of hierarchical keyword bugs in Lightroom (or was it Aperture) many, many years ago.  I found I had fewer issues and more flexibility doing it my hand.

    Example, I might want my cat to be listed under Family instead of the generic Animals.

    0
  • SFA

    Harry25,

    If you have created hierarchical keywords how would you want them to be deployed if not with the higher levels entered automatically?

    There is also some potential for the same KW to be required in multiple "Library" groups and for the Library group source to be identifiable.

    However once applied the principle is that an accepted KW, added to the metadata, also has a life of its own independently of how it became associated. That is very helpful in terms of search functions.

     

    Grant

     

    0
  • Harry25

    SFA,

    Maybe I wasn't clear: it's great that the higher levels are entered automatically.

    But I find the behaviour is inconsistent: when I move the keyword to a new hierarchy, the images that have the keyword get the the old AND the new hierarchy.

    If the new hierarchy didn't get applied I could understand it.

    If only the old hierarchy was removed and the new one applied I could also understand it (it would also be the expected behaviour for me)

    But this behaviour I find confusing and I think it's probably a bug, not a feature.

    0
  • Eric Valk

    Basically it comes to this: Moving a child keyword from one hierachy to another causes all images which have this keyword to still have the previous parents of the keyword (as well as the new parents)

    This principle applies to both keyword moves in your example, so I think it is consistent, not inconsistent.

    I don't see this as a bug, more as a simplistic design choice.

    The other design choice would be that the images which have the keyword lose the previous parents of the keyword which is moved. However this cannot be so if there are other Child keywords of the same parents.

    E.G.  you have an image with both your dog Felix and your cat Snoopy in it. You assign keywords Animals/Dogs/Felix  and Animals/Cat/Snoopy. You move Felix to the Family hierarchy. Now Capture One has to notice that there are other child keywords, Snoopy and Cat with the same parents assigned to image. So Animals must stay but Dogs has to go.

    Unless the user has made an error in keyword assignment, the old parent keywords are still relevant, so its not obvious that every user would want this improvement. Some users may already have work flows and scripts based on the current  behaviour.

    This behaviour as seen by a user is more complicated, because the results are different depending on other keywords which may be assigned to the image.

    Capture One SW would have to check the entire keyword hierarchy of every image with the keyword, to determine if the key word that is being moved is the only child of each of its parent keywords. This will slow down the performance somewhat.

    I think this change is a feature request, and it likely needs a user setting to control if parent keywords are deleted after a child keyword is deleteed or removed.

    0
  • Harry25

    I see what you mean, and that's exactly what I would expect the software to do.

    The keyword moving takes forever anyway for some strange reason (I only have about 30.000 images in my library) on quite beefed up MacBook Pro. 

    Even if it has to take some more time, it really doesn't matter in order to have better consistency in the catalog. 

    This is not an operation you should be doing often anyways :)

    In any case a user setting like you suggested would be nice or at worse a simple warning would be nice.

    I have opened up a support request for the same matter - I will post if and when I get a reply (I'm not very hopeful - previous attempts have taken weeks to get a boilerplate answer, and after a few weeks of pointless exchanges I've just given up. )

    0
  • Harry25

    Just to prove my point, and I guess everyone else's in this forum:

    After 8 days, I finally received a boiler-plate response with generic and 100% unrelated to my case questions (eg. test other sessions/catalogs, does my problem concur when viewing all raws or a particular one, etc).

    Let's see when I will get dignified with a proper response from them....

    0

Post is closed for comments.