Propagation of categories

Started by tvi55, May 12, 2014, 05:02:38 PM

Previous topic - Next topic

tvi55

I did some first testing of files which are connected with the version concept.

During the propagation of categories I discovered the following, which I would consider a bug as it might impact the introduction of versions in an existing database (eg migrated from IM3).

Observation:

If the target file includes categories which are not available in the master file, then they still exist after data propagation. They should however be removed to get the list of categories in sync. Such removal however needs to be limited to categories from the range of "categories to propagate" as defined in the "File Relation" preferences.

Mario

This should be discussed and maybe moved into a feature request.
I think that many people rely on existing categories not being removed in the versions. Consider keywords. Consider categories you only use for JPEG files but never for RAW - most likely you want to keep these categories even if you propagate.

I'll move this to General Discussions to other users can comment.
-- Mario
IMatch Developer
Forum Administrator
http://www.photools.com  -  Contact & Support - Follow me on 𝕏 - Like photools.com on Facebook

ubacher

Since we can control category propagation down to the individual category I consider tvi55's request as valid.

It is a question of definition: What does propagating a category mean:  Copy if set
or: Make it the same as master.

tvi55

If I may re-emphasize on my request ...

I would definitively consider a new or removed category within a category tree as a change within that tree which needs to be fully reflected in the other versions, provided that I included that tree in the "categories to propagate" as defined in the "File Relation" preferences.

In my situation I frequently refine my category "Location" by adding more levels (= details) to existing photos. The propagation feature would be completely useless to me if such changes would not be reflected in the versions.

For the situation mentioned - categories which are only used for JPEG files but never for RAW - such categories should never be included in the "categories to propagate", as they would be unique to a specific version only.

To summarize: categories from the "categories to propagate" as defined in the "File Relation" preferences should be treated as master during propagation, which overwrite whatever exists in the target files within the same domain.

Disclaimer: I explicitly exclude "keywords" from my request.

Ferdinand

Quote from: tvi55 on May 13, 2014, 08:50:06 AM
For the situation mentioned - categories which are only used for JPEG files but never for RAW - such categories should never be included in the "categories to propagate", as they would be unique to a specific version only.

I'm glad you put this qualification in, because otherwise I would be protesting loudly.

I understand what you are asking and why, but having written propagation scripts, while possible this is not so easy to do and adds complexity.  If it's always just the "Location" category perhaps you could run a script or something to strip that category from the versions before you update the master's category and trigger propagation.

Mario

As I said above, make a feature request with detailed info. I will look into this for a later version.
-- Mario
IMatch Developer
Forum Administrator
http://www.photools.com  -  Contact & Support - Follow me on 𝕏 - Like photools.com on Facebook

tvi55

Quote from: Ferdinand on May 13, 2014, 09:11:00 AM
If it's always just the "Location" category perhaps you could run a script or something to strip that category from the versions before you update the master's category and trigger propagation.
If have done this for years in IM3 using your script and a home grown one to keep the categories between my RAWs and JPGs in sync. I had expected (hoped) that I could get rid of these using IM5's version capability.

I will raise a feature request suggesting a logic similar to what you implemented in your script for IPTC data (see attachment) with the "clear/keep" option on the target files. This would then support both worlds of thoughts on this subject.

[attachment deleted by admin]

Ferdinand

Quote from: tvi55 on May 13, 2014, 11:15:08 AM
I had expected (hoped) that I could get rid of these using IM5's version capability.
Me too, but I am less optimistic at the moment.  See for example the thread about orientation propagation.