Quote:
Originally Posted by jhowell
I based that on post #355 in this thread by angryAspect which has files extracted from a Scribe that resulted from sending an EPUB and a PDF. The EPUB was converted to KFX reflowable format and the PDF was converted to KFX print replica format.
I also base it on having compared the Scribe firmware with the firmware used by other Kindles. The only sign of handwritten annotation support I found was in the part of the firmware that handles KFX format.
It would be good to have more data on how this works. I suggest that you connect your Scribe to a computer via USB and see what format actually results on the device from PDFs you send.
KFX print replica works similarly, but not quite the same, as that book.
In KFX print replica each page is rendered from a PDF file embedded within the KFX document. However the PDF is only used as a source of page images. Other functionality, such as links and text for search and dictionary lookup, are produced by adding an invisible fixed layout overlay on top of the page image. I am wondering whether handling that overlay is not yet working properly on the Scribe.
Or perhaps it depends on how the original PDF was structured. Maybe the conversion process cannot extract the text from some PDFs.
Possible, but highly unlikely in my opinion. |
I did some more edits to my post in light of other discoveries. Now I do recall that earlier post, so you're undoubtedly correct. My optimism for Print Replica support is restored, even as my hopes for improved fixed layer functionality is dashed. In any case a serious regression from PDF functionality.
And the current Export to PDF isn't going to impress anyone either.
I looked at what is on device. Everything has .kfx extension of course. And I don't see any obvious differences in the associated .sdr folders. So although I know which are converted from PDF and which are converted from ePub, because I sent them, it's not obvious from directory listing which is which.
I tried some other PDFs, none of them have selectable text on Scribe. But I'd guess the issue is upstream, not with Scribe itself.
And of course anyone reading this who has a Scribe, can report whether they can select text in the imported PDFs they are looking at. My prior is that one cannot.