[digiKam-users] Digikam, Tags and Raw files

[Matthieu]

Dear all,

I use Digikam for many years: this is a powerful tool which cover almost all the features I expect from this kind of software; bugs are also quickly fixed, new features are add very often.

But unfortunately, since tags are not written in the RAW files, I spend many time to manage my photo collection of 25k files: I have more than 66% of my photos which are raw files, the 33 remaining % are shared in JPG, and videos files.

Yesterday, for an unknown reason, my tags tree crashed.

I thought reloading tags from the metadatas stored in my files, but because they are not written anymore in the RAW files, the only process I have to restore my job is to do manually a second time.
Many months of job to redo.

How to backup digikam is not the subject of this email :)

I can understand that there are some risk to update RAW files.

But I'm ready to take this risk for the following reasons:
- For the last 2 or 3 years I updated my RAW files with both Digikam (when this feature was available), and with exiftool. I never encountered any issue.
- I use Digikam to help me managing my photo collection, not as a tool to view my picture collection: if I define a tag in a raw file in Digikam, I expect being able to view this tag in a different software (rawtherapee, lightraw, etc)
- Today, I have some differences between data stored in my RAW files content, and the Digikam database.

This is your product, with their own constraints. I use it as it because it was helping me to earn time.
But sometimes, I need to find some workarounds to adapt it as the way I manage my pictures.
Since this feature has been disabled, I was thinking editing correctly all my files in Digikam, and then develop a software / plugin / shell scripts / which would process the Digikam database and store the missing tags with exiftool.

The crash I got yesterday let me think differently...
So my question now: the checkbox to use the experimental feature to write tags in raw files is unchecked since the 5.9 release (those one I use). If I'm able to uncheck it in the sources files, will I be able to restore this feature? Or did you also remove all the writing process behind when disabling this feature?

If this is the second answer, which risks I 'll encounter to downgrade my Diigikam release to the latest one where this feature was available ? (5.8 or 5.7 if I remember well) ?

Thanks in advance for your help

Kind regards

Matthieu

[Gilles Caulier-4]

2018-07-29 5:56 GMT+02:00 Matthieu <[hidden email]>:

Dear all,

I use Digikam for many years: this is a powerful tool which cover almost all the features I expect from this kind of software; bugs are also quickly fixed, new features are add very often.

But unfortunately, since tags are not written in the RAW files, I spend many time to manage my photo collection of 25k files: I have more than 66% of my photos which are raw files, the 33 remaining % are shared in JPG, and videos files.

Yesterday, for an unknown reason, my tags tree crashed.

I thought reloading tags from the metadatas stored in my files, but because they are not written anymore in the RAW files, the only process I have to restore my job is to do manually a second time.
Many months of job to redo.

How to backup digikam is not the subject of this email :)

I can understand that there are some risk to update RAW files.

But I'm ready to take this risk for the following reasons:
- For the last 2 or 3 years I updated my RAW files with both Digikam (when this feature was available), and with exiftool. I never encountered any issue.
- I use Digikam to help me managing my photo collection, not as a tool to view my picture collection: if I define a tag in a raw file in Digikam, I expect being able to view this tag in a different software (rawtherapee, lightraw, etc)
- Today, I have some differences between data stored in my RAW files content, and the Digikam database.

This is your product, with their own constraints. I use it as it because it was helping me to earn time.
But sometimes, I need to find some workarounds to adapt it as the way I manage my pictures.
Since this feature has been disabled, I was thinking editing correctly all my files in Digikam, and then develop a software / plugin / shell scripts / which would process the Digikam database and store the missing tags with exiftool.

The crash I got yesterday let me think differently...
So my question now: the checkbox to use the experimental feature to write tags in raw files is unchecked since the 5.9 release (those one I use). If I'm able to uncheck it in the sources files, will I be able to restore this feature?

yes. return "true" to this method at line 144, and you will restore the feature :

https://cgit.kde.org/digikam.git/tree/core/libs/dmetadata/metaengine.cpp#n144

 

Or did you also remove all the writing process behind when disabling this feature?

Code is not removed.

In a better world, Exiv2 must have a check internally to see if current raw file processed is well supported at metadata writting process. In fact, to be right, Exiv2 must have a generic method to check is current type mime (and more info for file processed) are well supported at write process to prevent data corruption. As i know, this method do not exist yet.

The method in DK source code is just a simple hack.

 

Gilles Caulier

[Matthieu]

Thanks Gilles

2018-07-29 17:10 GMT+12:00 Gilles Caulier <[hidden email]>:

2018-07-29 5:56 GMT+02:00 Matthieu <[hidden email]>:

Dear all,

I use Digikam for many years: this is a powerful tool which cover almost all the features I expect from this kind of software; bugs are also quickly fixed, new features are add very often.

But unfortunately, since tags are not written in the RAW files, I spend many time to manage my photo collection of 25k files: I have more than 66% of my photos which are raw files, the 33 remaining % are shared in JPG, and videos files.

Yesterday, for an unknown reason, my tags tree crashed.

I thought reloading tags from the metadatas stored in my files, but because they are not written anymore in the RAW files, the only process I have to restore my job is to do manually a second time.
Many months of job to redo.

How to backup digikam is not the subject of this email :)

I can understand that there are some risk to update RAW files.

But I'm ready to take this risk for the following reasons:
- For the last 2 or 3 years I updated my RAW files with both Digikam (when this feature was available), and with exiftool. I never encountered any issue.
- I use Digikam to help me managing my photo collection, not as a tool to view my picture collection: if I define a tag in a raw file in Digikam, I expect being able to view this tag in a different software (rawtherapee, lightraw, etc)
- Today, I have some differences between data stored in my RAW files content, and the Digikam database.

This is your product, with their own constraints. I use it as it because it was helping me to earn time.
But sometimes, I need to find some workarounds to adapt it as the way I manage my pictures.
Since this feature has been disabled, I was thinking editing correctly all my files in Digikam, and then develop a software / plugin / shell scripts / which would process the Digikam database and store the missing tags with exiftool.

The crash I got yesterday let me think differently...
So my question now: the checkbox to use the experimental feature to write tags in raw files is unchecked since the 5.9 release (those one I use). If I'm able to uncheck it in the sources files, will I be able to restore this feature?

yes. return "true" to this method at line 144, and you will restore the feature :

https://cgit.kde.org/digikam.git/tree/core/libs/dmetadata/metaengine.cpp#n144

 

Or did you also remove all the writing process behind when disabling this feature?

Code is not removed.

In a better world, Exiv2 must have a check internally to see if current raw file processed is well supported at metadata writting process. In fact, to be right, Exiv2 must have a generic method to check is current type mime (and more info for file processed) are well supported at write process to prevent data corruption. As i know, this method do not exist yet.

The method in DK source code is just a simple hack.

 

Gilles Caulier