[davidfor]

Originally Posted by Rev. Bob

If you'll recall, that was my original theory: that the firmware wasn't properly handling database changes for my usage pattern, which primarily consists of adding books directly to the card without involving the device's USB cable.

Which is a usage pattern that a lot of other people have. The database is almost definitely broken. The question is whether this is related to this usage pattern or something else. If the problem reoccurs after a while with the new database, then it seems likely it is. But, if it doesn't, then it probably was something else.

Because (a) sometimes my device doesn't recognize a card insertion while it's turned on, and (b) I turn my device off when I'm not reading anyway.

OK. Other than a card that I have that was never recognised, the only times that I have had problems were when I slipped while inserting the card.

It started around 44% and increased in increments of a few points - pretty much as I'd expect, except for the length of time involved.

That is a bit different. My experience is that the percentage starts at 9% and then jumps in roughly equal steps, one for each book. When I put the card in with 500 books earlier, it counted by ones. I just added 10 books and it climbed by about 9% each time.

It I read what you are saying correctly, you are seeing a lot more and smaller steps that I would expect for six books. That suggest some books are being added each time you start the device. That is another sign of a corrupt database. Changes made while using the device seem OK, but can't be written to the database correctly. When you power off, the changes are lost and the restart reads the unchanged database.

If the firmware isn't properly maintaining the database, I'd consider that a serious firmware bug that should be fixed. I don't use sync software, I don't manually tinker with the database, I just copy books onto the card...so if the database is getting corrupted, what's to blame besides the firmware? And if you have to destroy the database and start from scratch every so often to avoid this kind of slowdown, how is that not a bug that needs to be fixed?

And I agree completely. My point was that this thread is about features people would like in the firmware that would improve the device. We are discussing a problem you are seeing on your device. The discussion would be much better in it's own thread.

The reader's back up now, and it did rescan the card much more quickly. I'm syncing over wifi now, and then I'll see if it remembers any shelving data. (You mentioned losing reading stats; you didn't mention shelves...which I put a lot more effort into. After all, I only read one book at a time, but I've gotta FIND the next one when I'm done!) Either way, my next step will be to reshelve the books (if necessary)

Any shelves that were on the device the last time you synced with Kobo, will be recreated. Any kepubs that were on shelves will be put back. Information about the sideloaded books are not synced to Kobo, so they can't automatically be put back onto shelves.

and then compare the new database to the old one. Even a basic size comparison should tell me something; the old one is just over 11 megs.

It depends on the problem. If the rows were being left in the database, then size will tell you something. If it was a corrupt table or index (my bet) then, the size will be roughly the same.

(And why do I have 600+ image sets in .kobo/images/ for books I've never even SEEN?)

They are probably for recommendations plus a there are some non-cover images in the folder.