I will only ever use indexing and will never, ever store anything in the dtbase file (please, I don’t want any commentary on this and why you like to put your stuff in the database, ok?–I get there are valid reasons for that preference. I have a different preference and my reasons are valid, OK?).
Everything I have ever indexed appears as indexed in the global inbox. Of course, the only way any document ever got indexed to begin with was that I indexed a folder in another database (e.g.–not the global index database).
So, every indexed entry in the global index database is a duplicate across the databases, not a duplicate within the global index database. In a word, there is NO reason to have these indexes twice and wast the 1.4G that the index takes.
Hence, how do I purge the global index? There is NO harm to do this as NO document is in any devonthink database. All files are in the file system (of course, they would be also if in the devonthink database, but in the screwball hidden file structure used) so no files can be lost. No indexes can be lost as they are all in the intended database.
Thanks. Sorry for the parenthetical commentary but it seems we indexers must defend ourselves from the ideological onslaught of the importers. Each to their own. There are good reasons that Devonthink accommodates both groups.
Guess that’s an onslaught I’ve never noticed. Half my stuff is indexed. Beautiful.
Anyway, you delete the indexed files then empty the trash. When you empty the DEVONthink trash, make sure you select the “only in database” option. If you select “files and folders” then the originals in the file system will be sent to the OS X trash. “Only in database” just removes all the metadata and pointers from the database(s).
Question: why was everything there to begin with? Indexing shouldn’t end up putting stuff in the inbox.
I’ll theorize that ages ago I did put everything IN, then indexed, but everything remained in the inbox–although that is not clear either. Since I would have, in the early exploration of DT, used the group names provided by the folder names.
Well, all is good. Now, let’s see what happens after a sync across machines.
I do wonder how it got that way but I think that is because in early experiments I did import all of the documents and folders, then chose to index them. So, the folders remained in the inbox. Though that is still a bit strange because I would have put the imported folders/documents into their proper database. A mystery…
