Help with PRC format

[cc_in_oh]

I'm still curious about why TOR would make the Kindle download a .prc instead of .mobi? From what I'm reading here it sounds like the mobi would be smaller, and they have a separate download for .epub...

[JSWolf]

Originally Posted by cc_in_oh

I'm still curious about why TOR would make the Kindle download a .prc instead of .mobi? From what I'm reading here it sounds like the mobi would be smaller, and they have a separate download for .epub...

What happened was Tor made the ePub and then used Kindlegen to make a dual Mobi/KF8 file. That's how they did it..mobi or .prc makes no difference. It's just a file extension.

[HarryT]

Originally Posted by cc_in_oh

I'm still curious about why TOR would make the Kindle download a .prc instead of .mobi? From what I'm reading here it sounds like the mobi would be smaller, and they have a separate download for .epub...

The extension has nothing whatsoever to do with the size of the file. What is it that you've read that makes you believe otherwise?

[pdurrant]

There's one instance in which KindleUnpack would extract an ePub that wasn't just the KF8 componenet repackaged. That would be if the file in question was to output from KindleGen and included the original source files, and those original source files were an ePub.

I've no idea if that's the case with the TOR freebies. I suspect not. With a .prc extension, perhaps they're still using the ancient Mobipocket Creator...

[HarryT]

Originally Posted by pdurrant

There's one instance in which KindleUnpack would extract an ePub that wasn't just the KF8 componenet repackaged. That would be if the file in question was to output from KindleGen and included the original source files, and those original source files were an ePub.

I've no idea if that's the case with the TOR freebies. I suspect not. With a .prc extension, perhaps they're still using the ancient Mobipocket Creator...

If KindleUnpack is producing an ePub file, doesn't that mean that it must either be a dual Mobi or (unlikely, as you say) that an ePub source file is included? Either of those two options mean that they must be using KindleGen to create it rather than Mobi Creator.

[pdurrant]

Originally Posted by HarryT

If KindleUnpack is producing an ePub file, doesn't that mean that it must either be a dual Mobi or (unlikely, as you say) that an ePub source file is included? Either of those two options mean that they must be using KindleGen to create it rather than Mobi Creator.

Good point. I don't think that KindleUnpack will (in the absence of other options) produce an 'ePub' from the Mobi 'HTML'.

[jhowell]

Examining the content of the MOBI format of TOR's free Spin by Robert Charles Wilson shows that Nate is correct.

The file shows that it was created by kindlegen V2.9 and it does contain an SRCS record with the source EPUB, identical to the EPUB that can be downloaded separately.

[cc_in_oh]

Originally Posted by HarryT

The extension has nothing whatsoever to do with the size of the file. What is it that you've read that makes you believe otherwise?

That the prc contains both mobi and epub, though presumably not a complete duplication of the text.

And that a previous freebie downloaded as a 556k epub and a 583k mobi, while this one was a 470k epub and a 1860k prc...

[pdurrant]

Originally Posted by jhowell

Examining the content of the MOBI format of TOR's free Spin by Robert Charles Wilson shows that Nate is correct.

The file shows that it was created by kindlegen V2.9 and it does contain an SRCS record with the source EPUB, identical to the EPUB that can be downloaded separately.

Thanks! That seems definitive.

Originally Posted by cc_in_oh

That the prc contains both mobi and epub, though presumably not a complete duplication of the text.

And that a previous freebie downloaded as a 556k epub and a 583k mobi, while this one was a 470k epub and a 1860k prc...

That sounds right, given jowell's finding. The "prc" contains THREE copies of the book text!

[HarryT]

Originally Posted by cc_in_oh

That the prc contains both mobi and epub, though presumably not a complete duplication of the text.

What TOR have done here is to make available for download the "raw" output from the Kindlegen tool, which contains both a dual Mobi (ie separate MOBI7 and KF8 versions of the book) and the source it was created from (the ePub). This is the file that a publisher provides to Amazon to go into the Kindle Store, but it really shouldn't ever be made available for download to anyone. Big mistake by someone at TOR. The fact that it has a ".prc" rather than a ".mobi" extension has no bearing on the matter whatsoever.

The comment that I made earlier in the thread - that the fact that the "KindleUnpack" Calibre plugin generates an ePub files does not mean that there actually is a separate ePub file - remains perfectly valid. You will never, ever download a file from Amazon that contains a separate ePub - the only reason that this one does is because the publisher has screwed up, basically. The normal action of KindleUnpack (for books that people actually download from Amazon) is to repackage the KF8 component of the book in an ePub wrapper.

[JSWolf]

Originally Posted by pdurrant

There's one instance in which KindleUnpack would extract an ePub that wasn't just the KF8 componenet repackaged. That would be if the file in question was to output from KindleGen and included the original source files, and those original source files were an ePub.

I've no idea if that's the case with the TOR freebies. I suspect not. With a .prc extension, perhaps they're still using the ancient Mobipocket Creator...

Using the KindleUnpack plugin, I was able to split the KF8 & Mobi parts. So no, it was not created using Mobipockt Creator. Looking at the OPF for the KF8, it has the following.

<!-- BEGIN INFORMATION ONLY 
<meta name="Cover ThumbNail Image" content="Images/image00345.jpeg" />
<meta name="ASIN" content="9c08395e-d12a-40ac-baeb-46345ccf9481" />
<meta name="Creator Software" content="202" />
<meta name="Creator Major Version" content="2" />
<meta name="Creator Minor Version" content="9" />
<meta name="Creator Build Number" content="0" />
<meta name="Font Signature (hex)" content="0300000000000000000000000000008000200000000000000000000000000000e5edbbfd01ae01f101c901a94094409340994098409c409d40a6" />
<meta name="StartOffset" content="10227" />
<meta name="Unknown_but_changes_with_file_name_only_542" content="8hD9" />
<meta name="Language_524" content="en" />
<meta name="K8(125)_Count_of_Resources_Fonts_Images" content="7" />
<meta name="K8(129)_Masthead/Cover_Image" content="kindle:embed:0006" />
<meta name="K8(131)_Unidentified_Count" content="5" />
<meta name="kindlegen_BuildRev_Number" content="0730-890adc2" />
<meta name="Has Fake Cover" content="0" />
<meta name="kindlegen_Input_Source_Type" content="kpr" />
<meta name="cdeType" content="EBOK" />
END INFORMATION ONLY -->

So this was definitely made from the ePub using Kindlegen.

[JSWolf]

The advantage of the way TOR is doing this is they give a KF8 for the Kindles that handle KF8 instead of a Mobi.