Hi
Yes Calibre is a great tool but unfortunately it still lucks lot of *small* things that may reduce satisfaction of anyone of us, so i used to solve them by writing my own mini programs, and come to a simple idea that helped me in that very much, maybe someone already passed through that, anyway I will put it here.
It will help you to do the following:
1- Dialing with Djvu files (that occupies 30% of my library) in calibre is a pain, it can't generate the covers from files nor "Pages count" plugin can extract pages count, this will solve both problems.
2- Writing Metadata to the files that are in the library itself instead of sending them to device to force Calibre do that.
3- I have a lot of OCRed, just scanned without OCR, or Native PDF files, and I usually need to know the type of the book (I'm using custom column for that), because reading OCRed books on portable devices is plainly slow and uncomfortable (no text wrap ect...) , by this method you can distinguish file contents.
4- Maybe other Ideas you can share with us
So how to do that? very easy, by any available Automation program, I tried the evolution version of WinAutomation that was more than enough to do all of those things, it will be not very fast, but leaving your library with it for a night was enough to fix those issues for thousand of my books!
Hope this will be Helpful!