View Full Version : opf to mobi to kindle freeze/reboot


tjcarter
06-26-2008, 06:51 AM
Hi all,

I have a .opf and friends that I tried to test out Igor's Kindle scripts and mobiperl on. The result seems to be a .mobi file, and my understanding is that this can go directly into the Kindle. I put it there, and sure enough the Kindle sees it, but it freezes when I try to open the file and then reboots after a bit.

I am wondering if there may be something in the .opf that wound up in the .mobi that breaks the Kindle, or if I'm just doing it wrong in general.

I haven't tried other .mobi files yet or to convert something from Project Gutenberg or anything like that.

Any thoughts?

tompe
06-26-2008, 07:16 AM
Do that happen for both the MobiPerl opf2mobi and for the other program?

Test the MobiPocket file with FBReader or the desktop reader to see that the file is basically OK.

HarryT
06-26-2008, 08:17 AM
If you're using a Windows PC, the other thing that would be worth doing would be to try creating the Mobi file with Mobi Creator (downloadable from the Mobi web site).

wallcraft
06-26-2008, 09:42 AM
I have a .opf and friends that I tried to test out Igor's Kindle scripts and mobiperl on. There should be no need to use Igor's scripts, because this ends up as a DRM-free MOBI file. If you are on a Windows desktop try using MobiPocket Creator (give it the .opf) or the command line version mobigen, both at Mobipocket Developer Center (http://www.mobipocket.com/dev/default.asp). There is also a Linux version of mobigen now. Usage: mobigen file.opf -c1 -s0

tjcarter
06-26-2008, 01:18 PM
I am a Mac user, so without the source to mobigen, the Linux version isn't going to be the easiest solution for me. The possibility exists I could set up a service like the Kindle email that will unpack the opf, do the conversion, and then email back the .mobi it creates. I suppose there's also virtualization of Windows.

I'll give the official tool a try on one of those platforms and see what results.

tompe
06-26-2008, 01:29 PM
I am a Mac user, so without the source to mobigen, the Linux version isn't going to be the easiest solution for me. The possibility exists I could set up a service like the Kindle email that will unpack the opf, do the conversion, and then email back the .mobi it creates. I suppose there's also virtualization of Windows.

I'll give the official tool a try on one of those platforms and see what results.

Was it MobiPerl that produced a file that did hang? If you give me the opf directory (or lit file) I an test opt2mobi and see if it works on my Gen3.

tjcarter
06-26-2008, 02:19 PM
Tommy, I'll drop you an email with it attached. It's not a public domain work, however I figure testing software with it must qualify as fair use, somehow. ;)

tompe
06-26-2008, 06:04 PM
Tommy, I'll drop you an email with it attached. It's not a public domain work, however I figure testing software with it must qualify as fair use, somehow. ;)

The file produced by opf2mobi did work in FBReader and on my Cybook Gen3. But I do not know if it is so useful since the TOC did not contain any links.

You could try to tun html2mobi on just the html file and see if this work on your Kindle.

wallcraft
06-26-2008, 07:35 PM
But I do not know if it is so useful since the TOC did not contain any links. This might be the problem. I have seen .opf files with no TOC guide item (they usually then have a tour which serves the same purpose), but I can't recall one with an empty TOC.

tjcarter
06-26-2008, 09:32 PM
Wallcraft, probably I should go back an re-post my intro message that didn't go through the first time, since it'll make a lot more sense then, but I'm converting a DAISY 3 text book (which should be OEBF compliant) for use on the Kindle. Navigation in DAISY books is somewhat unusual in that we typically navigate by heading level or by page using markup, and then by smaller increments using text features.

Typically, H1 marks chapters, H2 and sometimes H3 mark subheadings. Pages are marked in a different way.

DAISY books are generally intended to be spoken either by a speech synthesizer, or as recorded audio books with a lot more navigation than you'd get from, say, Audible.com books. It's possible to have a combination text and audio book (which is great for people with dyslexia), but they're not very common.

DAISY isn't commonly used other than by those with disabilities, though as of DAISY version 3, as I said, a DAISY book should be OEBF compliant.

I'll try the HTML. Maybe there's a good way to programmatically generate the TOC from the heading level navigation used by the books? Just a thought.

wallcraft
06-26-2008, 10:45 PM
I'll try the HTML. Maybe there's a good way to programmatically generate the TOC from the heading level navigation used by the books? MobiPocket Creator can generate a TOC from the headings, see Table of Contents Wizard (http://www.mobipocket.com/dev/article.asp?BaseFolder=creatorpublisher&File=toc_wizard.htm). This does not help on a Mac though.

HarryT
06-27-2008, 02:43 AM
Maybe there's a good way to programmatically generate the TOC from the heading level navigation used by the books? Just a thought.

Book Designer will do this. Works very well indeed. Again, though, no help if you're using a Mac, I'm afraid.

tompe
06-28-2008, 05:55 AM
I'll try the HTML. Maybe there's a good way to programmatically generate the TOC from the heading level navigation used by the books? Just a thought.

I added a ticket for daisy3 book on the MobiPerl trac so I do not forget this. Extending html2mobi to generate a TOC would be one solution I suppose.

I think I understand how the ncx and smil file works and the current opf2mobi will not handle this kind of structure.

tjcarter
07-05-2008, 12:24 PM
Thanks Tommy, as complex as this mess is, supposedly it's completely valid (or soon shall be). Unfortunately, this is what happens when you let W3C people get involved with something. DAISY predated OEBF, and was more or less its own thing (and dramatically simpler) in DAISY 2. Then someone decided that DAISY and OEBF should merge, since they're both books right?, and suddenly OEBF is supposed to be more than just an extended web browser, it's now a full multimedia-enabled extended web browser!

Consequently, many DAISY 2 playback softwares, even those capable of text-to-speech were not updated for DAISY 3, because a properly compliant DAISY 3 reader is more ambitious a project than writing a web browser, unless you can just shove a whole web browser core into your code and start extending it, of course. Those that exist aren't properly compliant and just ignore outright any tag they don't understand.

But that's a whole 'nother rant. I remain convinced that e-books should be an HTML subset or a DTD similar to an HTML subset with obvious book tags added, not an HTML superset.

Nobody cares what I think though. ;)

Anyway, I'm looking forward to some future opf2mobi that can handle this madness. :) Thanks again!