Is there some easy way (i.e, without use of Terminal) to set an interval for indexing or to simply pause it?
What are you trying to achieve?
No.
And I second @chrillek’s question.
Let me explain. I had a major system failure in which I lost current copies of some important data. So I started scrambling around looking for the most current backups among my three backup systems. The best source should have been Time Machine. Unfortunately, it had recently become glitchy, and it couldn’t restore anything in the normal way. But the disk still had a lot of data, so all I would need is a disk index to find and recover one or two critical files. DTP can index disks, so I tried it. The trouble with that indexing still never stopped for more than a few seconds even after the index seemed complete. That was a pain. Thus, the question here. I’m happy to report that the index was good enough, and I recovered what I needed.
Good for you. I don’t think that DT’s index was ever meant to be applied to a whole disk. Which is also kind of recursive if the database is located on the same disk: you index, the metadata are written to the database, which changes the content of the disk and requires another index run…
1 Like
The Time Machine disk was remote. And since DTP was on the local disk, I assume that the index was too.
Also, is DTP really not designed to index a whole disk? The warnings are ambiguous about whether the caution is about volume or the nature of the data, i.e, “uncurated.”
Does it really make sense do have DT index a whole disk, especially the one itself is installed on? What, for example, would you expect to gain from indexing apps or other arbitrary binary files?
1 Like
No, it does not make sense. But in the frenzied moments following the meltdown, it seemed to.
No, it is not and this is documented. DEVONthink is not a Finder nor a Spotlight replacement. It’s also not a duplicate file finder. You should not drop home directories or entire hard drives into a database.
2 Likes
The documentation should be clarified. If I had read what you’ve written here, I probably wouldn’t have indexed the whole disk. But I saw that word “uncurated,” which suggested to me a concern that didn’t apply in my case. None if the data was unknown, in from the wild, whatever. I was in a squeeze, and I saw a loophole. Close the loophole. Clarify the concern.
The manual’s chapter on “Importing and Indexing” has this at its end:
While DEVONthink can handle large amounts of data, it is inadvisable to import or index uncurated data, like entire hard drives or your Home directory. You should be judicious in what you put into your databases.
Emphasis mine.
It’s understandable that you were in a panic, but I also have the impression that you assumed that “indexing” in DT is meant to do something that it does not do. “Indexing” in DT does not create an index of your hard disk (or even parts of that): It adds the textual content and the metadata of the files you “index” to a database. That’s not what you were after, I suppose.
And there is, as I tried to mention before, a logical conundrum: If you try to index the disk DT is stored on, you send the program into an endless recursion. It will add the first file it sees to the database’s index. Which then, consequently, is modified. As the database is a folder on disk, DT must reindex this folder again. Which leads to a modification of said folder …
Another reason not to index a disk.
5 Likes
Yes, that clarifies it, as does the paragraph after. Thanks.
1 Like