I have a little over 3,200 annotation files that can be sorted in date order because each file has a YYYY/MM/DD preface. I’ve come across more than a few files that are always 174 bytes and are empty and also multiple instances of the same file. In those files, the page links in the file no longer work. I’ve had to recreate those files.
Is there a way to verify the integrity of the Annotation group? Since the files appear in the Annotation group after they are created in the “home” groups , there is a link there…
FWIW, I’ve also experienced an instance where, once created, the file did not appear in the Annotation group. Recreating the annotation file resulted in the same thing: it does not appear in the Annotation group. Do I perhaps have some kind of corruption issue that I need to investigate?
I run the Verify & Repair Database routine on a regular basis - at least once a week, sometimes more, depending on how much work I’m doing. It passes almost every time and if it doesn’t, I fix what it reports that needs attention.
Do I perhaps have some kind of corruption issue that I need to investigate?
Database “corruption” is not a common thing.
FWIW, I’ve also experienced an instance where, once created, the file did not appear in the Annotation group.
“an instance” is certainly not indicative of “corruption” or any longstanding issue. If it’s recurrent, especially consistently reproducible, that says something and is more useful to us.
Are you indexing?
What is Settings > Sync > Conflicts set to in DEVONthink or DEVONthink To Go?
I don’t index. I’m 100% imported. I also don’t synchronize with any other device.
Settings>Sync>Conflicts is set to the default, Automatic, 16 connections and Duplicate Documents.
While I agree with you that one instance is not evidence of corruption or a long-standing issue, it concerns me because the whole point of my creating these annotation files is to make them searchable. ABBYY does a creditable job but Gemini does much better. I also catch at least as many errors by the newspapers as those generated by Gemini.
If an annotation file is not in the Annotation group, it obviously cannot be searched and this defeats the purpose of creating the file in the first place. Not all keywords show up in an OCR search. And, of course, a name like Smith that OCR transcribes as Zxu?w is not going to be found at all.
My search scope changes, depending on the circumstances, obviously. In my observation, I’m only targeting the Annotations group. That is because OCR might “see” Baker as Rokan. The name is spelled correctly in the Gemini transcription, however. If that particular file is not in the Annotation group, that occurrence of ‘Baker’ won’t be found.
In the group I’m working in, I have a total of 3,971 files, of which 3,041 have been processed. The annotation group, however, has 3,282 files, an overage of 241 files. To me, that indicates a problem that needs to be looked into.
The Annotations group isn’t incapable of containing non-Annotation files. Annotation files will have property icons for incoming and outgoing links. Look for ones that don’t have them.
Also, if you don’t have the option enabled to move them automatically, referring files moved to other databases would still be linked but not in the same database.