[geek1011]

Originally Posted by davidfor

The problem with that is that it is rare for others to be talking about the full conversion. Nearly all the discussion about sending kepub to Kobo devices is about the KoboTouchExtended driver. And, as I said before, I find it hard to believe that you didn't know it didn't do a full conversion this based on how many times it has been discussed including me explicitly mentioning it in the thread you started about developing a plugin.

I don't remember any explicit mentions of the difference between the conversions. I honestly thought they were pretty much the same thing. I'll do some more testing (about the differences) next time I use Calibre (which will probably be in a few weeks, as I want to try a few things with MOBI and AZW files, which don't have good standalone tools, and I'm not interested in making them).

Which really means that you want to do it all yourself, so you feel that calibre is intrusive. Whereas a lot of people are just glad there is something to help them.

I never said KTE was bad, I only said that there are benefits and uses for both. Calibre is intrusive for my workflow, as in, I don't use Calibre itself, and I also automate the syncing of books between my devices (and KTE doesn't work from the command line). Also, Calibre just isn't suitable for running directly on the Kobo (which I have been experimenting with recently).

My feeling is that doing the cleanup of the code during the send to the device is not a good idea. You aren't really pushing kepubify as just that, but, having kepubs lying around otherwise is fairly pointless.

I do keep kepubs lying around for various kepub-related experiments. Sometimes, I also keep them lying around so I can quickly transfer them and play around with them. And it is optional to keep them lying around. Also, when else would I do the cleanup? Another reason for the cleanup to happen during the conversion is that not all readers are sensitive to malformed HTML in the same way (some other readers parse it as XML and fail on the slightest syntax error, but are fine with self-closing title tags), and the cleanup kepubify does is specifically intended for the kepub reader. Quite a few people like their original books untouched.

Something like that would have a lot of code in it and a conversion would probably be a problem. I'd be interested to see if it survived the KoboTouchExtended driver.

The last time I tried converting it with a full conversion (and actually waited for it to finish), it mostly survived the process, but the fonts on the code were a bit messed up. Surprisingly, the spacing worked fine.

Somewhere I think I suggested using the calibre library maintenance to find and add books sent added to the library directories in this way.

Yes, I think I remember that too. I've actually been getting more requests for how to do this recently, so I'm considering adding it as a separate command for kepubify. How stable is the format of Calibre's metadata database?